Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs

Article

Fatty acid chain length drives

lysophosphatidylserine-dependent immunological

outputs

Highlights

Authors

d

Synthesis of methyl esters of lyso-PSs with varying lipid tails

and their metabolism

Neha Khandelwal, Minhaj Shaikh,

Amol Mhetre, ...,

Kithiganahalli Narayanaswamy Balaji,

Harinath Chakrapani,

Very-long-chain lyso-PSs signal through TLR2 to cause

neuroinﬂammation

Siddhesh S. Kamat

Role of long-chain lyso-PSs in macrophage signaling and

mast cell degranulation

Correspondence

amol@iiserpune.ac.in (A.M.),

harinath@iiserpune.ac.in (H.C.),

siddhesh@iiserpune.ac.in (S.S.K.)

ABHD12 is a major lyso-PS lipase in primary mast cells

In Brief

Khandelwal et al. synthesize a library of

methyl esters of lysophosphatidylserines

(lyso-PSs) to facilitate rigorous structure-

activity relationship studies, toward

understanding the role played by the lipid

tail of lyso-PSs in regulating the activation

of macrophages and mast cell

degranulation.

Khandelwal et al., 2021, Cell Chemical Biology 28, 1–11

June 17, 2021 ª 2021 Elsevier Ltd.

https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Fatty acid chain length drives

lysophosphatidylserine-dependent

immunological outputs

Neha Khandelwal,^1,5Minhaj Shaikh,^2,5Amol Mhetre,^1,5,Shubham Singh,^1,5Theja Sajeevan,¹Alaumy Joshi,^1,4

*

1 ,6 ,

Kithiganahalli Narayanaswamy Balaji,³Harinath Chakrapani,^2,and Siddhesh S. Kamat

*

¹Department of Biology, Indian Institute of Science Education and Research (IISER) Pune, Dr. Homi Bhabha Road, Pashan, Pune,

Maharashtra 411008, India

²Department of Chemistry, Indian Institute of Science Education and Research (IISER) Pune, Dr. Homi Bhabha Road, Pashan, Pune,

Maharashtra 411008, India

³Department of Microbiology and Cell Biology, Indian Institute of Science (IISc), Bangalore, Karnataka 560012, India

⁴Present address: Department of Biochemistry & Biophysics, Texas A&M University, College Station, TX 77843, USA

⁵These authors contributed equally

⁶Lead contact

*Correspondence: amol@iiserpune.ac.in (A.M.), harinath@iiserpune.ac.in (H.C.), siddhesh@iiserpune.ac.in (S.S.K.)

https://doi.org/10.1016/j.chembiol.2021.01.008

SUMMARY

In humans, lysophosphatidylserines (lyso-PSs) are potent lipid regulators of important immunological pro-

cesses. Given their structural diversity and commercial paucity, here we report the synthesis of methyl esters

of lyso-PS (Me-lyso-PSs) containing medium- to very-long-chain (VLC) lipid tails. We show that Me-lyso-PSs

are excellent substrates for the lyso-PS lipase ABHD12, and that these synthetic lipids are acted upon by

cellular carboxylesterases to produce lyso-PSs. Next, in macrophages we demonstrate that VLC lyso-PSs

orchestrate pro-inﬂammatory responses and in turn neuroinﬂammation via a Toll-like receptor 2 (TLR2)-

dependent pathway. We also show that long-chain (LC) lyso-PSs robustly induce intracellular cyclic AMP

production, cytosolic calcium inﬂux, and phosphorylation of the nodal extracellular signal-regulated kinase

to regulate macrophage activation via a TLR2-independent pathway. Finally, we report that LC lyso-PSs

potently elicit histamine release during the mast cell degranulation process, and that ABHD12 is the major

lyso-PS lipase in these immune cells.

INTRODUCTION

potent bioactivities in the mammalian central nervous and im-

mune system.

Cellular pharmacological studies have shown that lyso-PSs

Lipids have long been known as potent signaling molecules that

mediate many important physiological processes in mammals, regulate several immunological processes (Shanbhag et al.,

including humans (Wymann and Schneiter, 2008; Dennis, 2020) such as macrophage activation to clear apoptotic cells

2016; Fahy et al., 2005). Prominent among the signaling lipids (Frasch and Bratton, 2012), mast cell degranulation (Lloret and

are the prostaglandins (Dennis and Norris, 2015), the endocan- Moreno, 1995), leukemic cell stimulation (Park et al., 2005),

nabinoids (2-arachidonoylglycerol [2-AG] and anandamide chemotaxis of human gliomas (Lee et al., 2008), and maturation

[AEA]) (Blankman and Cravatt, 2013; Fowler et al., 2005), and of regulatory T cells (Barnes et al., 2015), and perhaps signal

the well-studied lysophospholipids, sphingosine 1-phosphate through Toll-like receptors (TLRs) (Van Der Kleij et al., 2002)

(S1P) (Gonzalez-Cabrera et al., 2014; Rosen et al., 2013) and ly- and/or G-protein-coupled receptors (GPCRs) (Inoue et al.,

sophosphatidic acid (lyso-PA) (Ishii et al., 2004; Contos et al., 2012) in the mammalian nervous and immune system. Interest-

2000). Given their physiological importance, the biosynthetic/ ingly, and of biomedical relevance, mutations to the putative

degradative enzyme(s) and/or cognate receptor(s) of the afore- lyso-PS receptors in humans have been linked to different auto-

mentioned lysophospholipids (S1P and lyso-PA) are pharmaco- immune diseases (Szymanski et al., 2014; Napier et al., 2015;

logical targets for drugs already in clinical use or under investiga- Chu et al., 2013). Murine studies have recently shown that accu-

tion in different phases of clinical trials for an array of human mulation of lyso-PS, especially very-long-chain (VLC) lyso-PSs

neurological and immunological disorders (Gardell et al., 2006; (Blankman et al., 2013), in the mammalian brain is a major cause

Yanagida and Valentine, 2020). Recently, the lysophosphatidyl- that drives the pathology of the early-onset human neurological

serines (lyso-PSs) have emerged as yet another important class disorder PHARC (polyneuropathy, hearing loss, ataxia, retinitis

of signaling lysophospholipids (Shanbhag et al., 2020), with pigmentosa, and cataract) (Fiskerstrand et al., 2009, 2010).

Cell Chemical Biology 28, 1–11, June 17, 2021 ª 2021 Elsevier Ltd.

1

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Figure 1. Structure of Me-lyso-PSs, and

testing them as substrates against ABHD12

(A and B) The chemical structures of (A) commer-

cially available canonical natural lyso-PSs and (B)

our synthetic Me-lyso-PS lipid library with both (R)-

and (S)-stereoisomers.

(C) Enzyme kinetic assays for membrane lysates

(10 mg) of HEK293T cells transfected with hABHD12

tested against the (R)-Me-lyso-PSs (0–400 mM,

30 min, 37^ꢀC, n = 3/data point). The line connecting

the points represents a ﬁt to the Michaelis-Menten

enzyme kinetics equation. See Table 1 for all enzyme

kinetics parameters.

(D) Lipase assays for (R)-Me-lyso-PSs (100 mM,

30 min, 37^ꢀC, n = 3/group) tested against brain

membrane lysates (20 mg) from wild-type (WT) (+/+)

or ABHD12 knockout (ꢁ/ꢁ) mice. All data presented

in (C) (data points) and (D) (bars) are presented as

mean standard deviation (SD).

to lyso-PSs. Of note, we report the distinct

contribution of individual lyso-PS lipids in

activation and pro-inﬂammatory responses

Interestingly, PHARC is caused by deleterious mutations to the from macrophages and in the release of histamine during mast

abhd12 gene (Fiskerstrand et al., 2009, 2010), which encodes cell degranulation, and speciﬁcally show how the fatty acid chain

an integral membrane metabolic serine hydrolase ABHD12 (a/b length of lyso-PSs plays a critical role in the output of the afore-

hydrolase domain-containing protein 12), a major and, to date, mentioned important immunological processes. Finally, we also

only in vivo functionally characterized lyso-PS lipase (Kamat annotate the metabolic serine hydrolase ABHD12 as a major

et al., 2015; Blankman et al., 2013; Singh et al., 2020).

lyso-PS lipase in primary mast cells, and expand the possible

Given the strong link between lyso-PSs and human diseases, physiological role of this biomedically important lipase to an

mechanistic studies are much needed to understand in more additional immunological process.

detail and biochemically characterize the signaling pathways

inﬂuenced by lyso-PSs in vivo. Two factors complicate and/or RESULTS

limit such studies: (1) unlike other signaling lipids (e.g., S1P, 2-

AG, AEA), in vivo lyso-PSs are found esteriﬁed with different fatty Synthesis of Me-lyso-PS lipids

acids (ranging from medium-chain [C10] to VLC [C24]) (Blank- Structurally, lyso-PSs are composed of three building blocks: (1) a

man et al., 2013; Barnes et al., 2015), and hence the precise sugar backbone (glycerol), (2) an amino acid head group (phos-

physiological contributions of the individual lyso-PSs remain pho-L-serine), and (3) a lipid tail (fatty acid) (Figure 1A) (Fahy

cryptic; (2) commercially, lyso-PSs are limited, and even those et al., 2011). It is this intricate combination of hydrophilic (glycerol

that are available are esteriﬁed only with long-chain (LC) fatty and phospho-L-serine) and hydrophobic (fatty acid) biological

acids (C16:0, C18:0, and C18:1) (Figure 1A). Therefore, delin- building blocks that confers amphiphilic properties to lyso-PSs

eating the biological contributions of medium or VLC lyso-PSs and enables them to access different cellular membranes, organ-

has not been possible to date.

elles, and compartments. Given this amphiphilic property, lyso-

To address the aforementioned problems, here we describe PSs serve as potent hormone-like mediators of various important

the synthesis of methyl esters of lyso-PS (Me-lyso-PS) bearing immunological processes (Vance, 2015; Tracey et al., 2018;

fatty acids of varying chain lengths ranging from medium-chain Shanbhag et al., 2020). From a stereochemical perspective, all

(C10–C14) to LC (C16–C20) to VLC (>C20) having the canonical natural lyso-PSs have two chiral centers: (1) a-carbon of the phos-

natural (R)- or unnatural (S)-conﬁguration at the glycerol back- pho-L-serine head group and (2) sn-2 carbon of the (R)-glycerol

bone. Next, we test these synthetic Me-lyso-PS lipids in a variety backbone (Figure 1A) (Mallik et al., 2018). Biosynthetically, lyso-

of biological assays, and show that the canonical (R)-Me-lyso- PSs are made from phosphatidylserine (PS) precursors, by the

PSs are highly bioactive while the corresponding unnatural (S)- enzymatic action of PS-speciﬁc phospholipases (Hosono et al.,

Me-lyso-PSs are largely inactive. Speciﬁcally, we show that 2001; Kamat et al., 2015), and subsequent studies have shown

(R)-Me-lyso-PS are excellent substrates for the PHARC-associ- that physiologically, almost all lyso-PSs exists as 1-(fatty acyl)-

ated lyso-PS lipase ABHD12, and conﬁrm this enzyme’s prefer- 2-hydroxy-sn-glycero-3-phospho-L-serine (Iwashita et al., 2009;

ence for VLC lyso-PS lipids (Joshi et al., 2018; Kamat et al., 2015; Kamat et al., 2015). Given the synthetic challenges in making

Blankman et al., 2013). Furthermore, we show that cellular car- this bioactive lipid, for example, (1) combination of the aforemen-

boxylesterases hydrolyze the methyl-ester moiety of Me-lyso- tioned hydrophilic and hydrophobic moieties, (2) lipophilicity of

PSs to yield lyso-PSs, and thus demonstrate that the synthetic the ﬁnal molecule, and (3) the need for retention in conﬁguration

Me-lyso-PSs serve as stable prodrug-like biological precursors of the two chiral centers toward making the correct diastereomer,

2

Cell Chemical Biology 28, 1–11, June 17, 2021

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Table 1. Kinetic constants for lyso-PS and Me-lyso-PS substrates tested against hABHD12 (HEK293T membrane lysates transiently

transfected with hABHD12)

Fatty acid chain

V_max/K_M(nmol/mg

protein/min/M) (3 10⁵)

Lyso-PS species

length: unsaturation

V_max(nmol/mg protein/min)

K_M(mM)

1a

16:0

18:0

18:1

10:0

12:0

14:0

16:0

18:0

20:0

22:0

24:0

10:0

12:0

14:0

16:0

18:0

20:0

22:0

24:0

15.6 0.6

18.3 0.5

17.8 0.6

2.9 0.2

7.6 0.3

12.2 0.6

15.2 0.6

17.2 0.6

19.0 0.6

22.0 0.6

22.9 0.6

0.2 0.02

0.4 0.05

0.6 0.1

0.8 0.1

1.2 0.1

1.3 0.2

1.2 0.2

1.2 0.3

72

48

46

8

4

6

2.2 0.3

1b

3.8 0.4

1c

3.8 0.5

(R)-2a

(R)-2b

(R)-2c

(R)-2d

(R)-2e

(R)-2f

(R)-2g

(R)-2h

(S)-2a

(S)-2b

(S)-2c

(S)-2d

(S)-2e

(S)-2f

(S)-2g

(S)-2h

126 18

131 14

106 14

0.2 0.04

0.6 0.06

1.1 0.1

67

47

43

40

7

5

4

2.3 0.3

3.7 0.5

4.4 0.5

5.5 0.6

5.7 0.6

275 32

255 35

215 28

187 24

178 25

165 23

167 26

176 31

0.007 0.001

0.016 0.002

0.028 0.003

0.043 0.006

0.067 0.008

0.078 0.009

0.072 0.009

0.068 0.008

only three naturally occurring LC lyso-PSs (1a–1c) are commer- etry (LC-MS)-based substrate assays (Joshi et al., 2018), we

cially available, thus limiting any rigorous structure-activity rela- found that both recombinant human ABHD12 (hABHD12) (Table

tionship (SAR) studies for this important lysophospholipid class 1 and Figure S1A) and endogenous mouse brain ABHD12

(Figure 1A). Since the commercial paucity of lyso-PSs would (mABHD12) (Figure S1B) robustly turned over (R)-Me-lyso-PSs,

impede our proposed studies, here we describe a facile synthesis and for the same fatty acid, LC (R)-Me-lyso-PS and correspond-

route toward making Me-lyso-PS esteriﬁed with saturated fatty ing canonical natural LC lyso-PS behaved almost identically.

acids of varying chain lengths ranging from medium-chain to Next, we performed rigorous enzyme kinetics measurements

VLC (Figure 1B and Data S1). Speculating that the free reactive for our (R)-Me-lyso-PS library against hABHD12 (Figure 1C and

carboxylate moiety of the amino acid end of the molecule might Table 1) and found that hABHD12 strongly prefers VLC (R)-Me-

cross-react and complicate the synthesis, we decided to make lyso-PSs as substrates, with (R)-2h (C24:0, V_max= 22.9

the Me-lyso-PS version. Our synthetic route successfully afforded 0.6 nmol/mg protein/min, K_M= 40 4 mM) and (R)-2g (C22:0,

the naturally occurring canonical (R)-Me-lyso-PSs (Figure 1B and V_max= 22.0 0.6 nmol/mg protein/min, K_M= 40 4 mM) being

Data S1) and the unnatural (S)-Me-lyso-PSs (Figure 1B and Data the best substrates. Furthermore, mouse brain membrane ly-

S1) with the same fatty acids (Figure 1B) to yield ﬁnal compounds sates were assayed against the same Me-lyso-PS library, and

((R)-2a–2h or (S)-2a–2h) in milligram quantities.

here too we found that mABHD12 strongly prefers VLC (R)-Me-

lyso-PSs, with (R)-2h (C24:0, rate = 8.2 0.3 nmol/mg protein/

min) and (R)-2g (C22:0, rate = 8.0 0.3 nmol/mg protein/min) be-

ing the best substrates (Figure 1D). In this experiment, mouse

brain membrane lysates from ABHD12-null mice were used as

controls to delineate speciﬁc contributions from mABHD12 (Fig-

ure 1D). In both these assays, we also tested (S)-Me-lyso-PSs

and found, not surprisingly, that these were very poor substrates

for hABHD12 (Table 1 and Figure S1C) and mABHD12 (Fig-

ure S1D), with catalytic efﬁciencies 100-fold lower than that of

corresponding (R)-Me-lyso-PSs for hABHD12 (Table 1). Taken

together, these substrate assays conclusively show that the

ABHD12-catalyzed lyso-PS lipase reaction is highly stereospe-

ciﬁc, prefers VLC (R)-lyso-PSs as substrates, and, together

with recent ﬁndings (Singh et al., 2020; Joshi et al., 2018), now

provides a concrete biochemical explanation as to why VLC

lyso-PSs accumulate the most in ABHD12 knockout mouse

brains (Blankman et al., 2013).

ABHD12 prefers VLC lyso-PSs as substrates

The lipids previously tested as substrates for the mammalian

lyso-PS lipase ABHD12 have been commercially available LC

lyso-PSs (1a–1c) (Blankman et al., 2013; Kamat et al., 2015) (Fig-

ure 1A) and various monoacylglycerol (MAG) lipids (Blankman

et al., 2007; Joshi et al., 2018; Navia-Paldanius et al., 2012). In

the absence of a commercial source for medium or VLC lyso-

PSs, which to the best of our knowledge have not been tested

against ABHD12, the MAG lipids have served as excellent surro-

gates for performing SAR studies toward biochemically under-

standing the substrate preference for this lipase (Joshi et al.,

2018). Therefore, having synthesized the library of (R)- and (S)-

Me-lyso-PSs containing medium-chain, LC, and VLC variants

(nomenclature in Figure 1B), we ﬁrst wanted to test whether

these lipids were indeed substrates for ABHD12. Leveraging es-

tablished liquid chromatography coupled with mass spectrom-

Cell Chemical Biology 28, 1–11, June 17, 2021

3

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Figure 2. Metabolism of Me-lyso-PSs by car-

boxylesterases and the pro-inﬂammatory ac-

tivity of Me-lyso-PSs in macrophages

(A) Top: the enzymatic reaction for metabolism (hy-

drolysis) of Me-lyso-PSs to the corresponding lyso-

PSs catalyzed by carboxylesterases. Bottom: ex-

tracted ion chromatograms from a LC-MS analysis

of (R)-2b (C12:0) ([M-H]^–= 454.221), (R)-2e (C18:0)

([M-H]^–= 538.315), and (R)-2h (C24:0) ([M-H]^–

=

622.409) showing the complete conversion of the

parent compound (Me-lyso-PS) to the correspond-

ing lyso-PS (loss of 14 Da) following treatment with

active (blue trace), but not denatured (red trace),

porcine liver carboxylesterase (0.1 U, 15 min, 37^ꢀC).

In these assays, 100 mg of the (R)-Me-lyso-PS was

used, and

a no-enzyme standard only control

(dotted black trace) was also included for the same

treatment. This LC-MS experiment was done twice

for each of the (R)-Me-lyso-PS, with reproducible

results each time.

(B) Secreted TNF-a and IL-6 from PPMs harvested

from WT (+/+) or ABHD12 knockout (ꢁ/ꢁ) mice

following treatment with vehicle (DMSO) or lipo-

polysaccharide (LPS) or (R)-2a–2h (1 mM, 4 h, 37^ꢀC);

All data are presented as mean

SD (n =

4/group). *p < 0.05, **p < 0.01, ***p < 0.001 versus

(+/+) group by Student’s two-tailed unpaired para-

metric t test.

Cellular carboxylesterases metabolize Me-lyso-PS to

lyso-PS

can hydrolyze the methyl group of (R)-Me-lyso-PSs to yield the

corresponding canonical (R)-lyso-PSs (Figure 2A) and that the

Having synthesized Me-lyso-PSs, we postulated that these might synthetic Me-lyso-PSs serve as stable prodrug-like biological sur-

be metabolized by cellular carboxylesterases to yield the corre- rogates to lyso-PSs, making them amenable to testing in primary

sponding lyso-PSs (Figure 2A). To test this hypothesis, we incu- macrophages and mast cells.

bated three (R)-Me-lyso-PSs of varying lipid tails ((R)-2b [C12:0,

medium-chain], (R)-2e [C18:0, LC], or (R)-2b [C24:0, VLC]) with VLC lyso-PSs elicit robust pro-inﬂammatory responses

active or denatured porcine liver carboxylesterase, and upon in mammalian macrophages

this treatment, by LC-MS (Joshi et al., 2018) we checked for the Lyso-PSs have previously been shown to elicit immunological

formation of the corresponding canonical lyso-PSs. We found responses via pro-inﬂammatory cytokine secretion (e.g., tumor

for all the three (R)-Me-lyso-PSs subjected to this treatment that necrosis factor a [TNF-a], interleukin-6 [IL-6]) from mammalian

upon incubation with active but not the denatured carboxylester- macrophages (Kamat et al., 2015; Frasch and Bratton, 2012),

ase, the (R)-Me-lyso-PS was completely consumed and a new and we wanted to assess whether the Me-lyso-PSs were capable

peak was observed with mass shift of ꢂ14 Da (parent m/z ꢁ 14 of the same, having shown that carboxylesterases in primary mac-

Da) (Figure 2A). The mass of this new peak corresponds to the rophages can metabolize Me-lyso-PSs to lyso-PSs (Figure S2A).

loss of the methyl group from (R)-Me-lyso-PS to yield the We found that when primary peritoneal macrophages (PPMs) iso-

corresponding canonical (R)-lyso-PS via hydrolysis by the carbox- lated from wild-type (WT) mice were treated with equal concentra-

ylesterase (Figure 2A). Since we planned to test this synthetic tions of the canonical LC lyso-PSs (1a [C16:0] and 1b [C18:0]) and

Me-lyso-PS library for different biological activities in mammalian the corresponding (R)-Me-lyso-PSs ((R)-2d and (R)-2e) of the

macrophages and mast cells, we decided to test whether lysates same lipid tail, they secreted almost equal amounts of pro-inﬂam-

of these immune cells had carboxylesterases capable of convert- matory cytokines (TNF-a, IL-6) (Figure S2B). Not surprisingly,

ing (R)-Me-lyso-PSs to the canonical lyso-PSs. To negate any treating PPMs with unnatural (S)-Me-lyso-PSs (Figure S2C), cor-

lyso-PS lipase type activity, we used active or denatured lysates responding free fatty acids (Figure S2C), or other lysophospholi-

from ABHD12-null primary peritoneal macrophages or perito- pids (lysophosphatidic acid [lyso-PA], lysophosphatidylglycerol

neal-derived cultured mast cells, and found that the active but [lyso-PG], and lysophosphatidylcholine [lyso-PC]) with same lipid

not denatured lysates from both these immune cells converted tail (Figure S2D) failed to elicit any inﬂammatory response, sug-

the three previously tested (R)-Me-lyso-PSs, namely, (R)-2b gesting that ligand recognition in PPMs in eliciting pro-inﬂamma-

(C12:0, medium-chain), (R)-2e (C18:0, LC), and (R)-2h (C24:0, tory responses is highly stereospeciﬁc for (R)-lyso-PSs. Having

VLC), to their corresponding canonical (R)-lyso-PS lipid (Fig- established comparable bioactivities of canonical lyso-PSs and

ure S2A). Our results clearly show that cellular carboxylesterases (R)-Me-lyso-PSs, we isolated PPMs from WT or ABHD12

4

Cell Chemical Biology 28, 1–11, June 17, 2021

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Figure 3. VLC lyso-PSs elicit pro-inﬂamma-

tory responses via

pathway

a

TLR2-dependent

(A and B) Secreted TNF-a and IL-6 from PPMs

harvested from: (A) WT mice pre-treated with DMSO

or SpaB (10 mM, 4 h, 37^ꢀC), followed by treatment

with vehicle or LPS or (R)-2a–2h (1 mM, 4 h, 37^ꢀC); (B)

WT (+/+) or TLR2 knockout (ꢁ/ꢁ) mice following

treatment with vehicle or LPS or (R)-2a–2h (1 mM,

4 h, 37^ꢀC).

(C) Representative images from Iba-1 immuno-

staining for microglial activation (scale bars, 250 mm)

and quantiﬁcation of enlarged cells (>200 mm²) in the

cerebellum (per 1.44 mm²) of WT (+/+) or TLR2

knockout (ꢁ/ꢁ) mice following intravenous injection

of vehicle (PBS), LPS, or (R)-2h (C24:0) (all 1 mg/kg

body weight, 10 h).

All data are presented as mean SD (n = 4–6/group).

macrophages (Figure 2B). We also per-

formed a similar SAR study in human

THP-1 macrophages and found similar re-

sults whereby VLC (R)-Me-lyso-PSs pro-

duced the highest secretion of pro-inﬂam-

matory cytokines (Figure S2E).

VLC lyso-PSs signal via TLR2 in

mammalian macrophages

Reported literature speculates that VLC

lyso-PSs perhaps signal through TLR2

(Van Der Kleij et al., 2002) although, to the

best of our knowledge, VLC lyso-PSs

have never been directly tested against

any receptor for any immunological activity

given their commercial unavailability. We

found from a large-scale gene expression

database (Figure S3A) (Wu et al., 2016)

knockout mice and incubated them with the (R)-Me-lyso-PS li- and by RT-PCR analysis (Figure S3B) that TLR2 (but none of

brary to determine whether the length of lipid tail had any effect the other putative lyso-PS receptors [Inoue et al., 2012]) is en-

on pro-inﬂammatory cytokine secretion. Interestingly, we found riched on WT PPMs. Next, we found that pharmacological

from this SAR study that VLC (R)-Me-lyso-PSs, particularly (R)- antagonism of TLR2 using Sparstolonin B (SpaB) (Liang et al.,

2g (C22:0) and (R)-2h (C24:0), produced the highest pro-inﬂam- 2011) ablated the increased pro-inﬂammatory cytokine secretion

matory cytokine secretion (TNF-a, IL-6) from WT PPMs caused by VLC (R)-Me-lyso-PS treatments in WT PPMs (Fig-

(Figure 2B). Not surprisingly, ABHD12-null PPMs secreted signif- ure 3A), and in human THP-1 macrophages (Figure S2F). Here,

icantly more pro-inﬂammatory cytokines compared with WT we found that SpaB treatment also ablated the increased pro-in-

PPMs upon (R)-Me-lyso-PS treatments, consistent with their ﬂammatory cytokine secretion caused by LPS treatment in WT

diminished lyso-PS lipase activity (Kamat et al., 2015) (Figure 2B). PPMs (Figure 3A), consistent with SpaB’s dual TLR2/TLR4

For either genotype, the pharmacological Me-lyso-PS treatments antagonism activity (Liang et al., 2011, 2013). Furthermore, we

did not have any effect on cell viability.

harvested PPMs from TLR2 knockout mice (Figure S3C) (Holla

Here, we used the bacterial outer membrane glycolipid lipo- et al., 2016) and measured pro-inﬂammatory cytokine secretion

polysaccharide (LPS) as a positive control, as this endotoxin following (R)-Me-lyso-PS treatments. Consistent with the phar-

tested in PPMs mimics Gram-negative bacteria and robustly macological studies, we found that VLC (R)-Me-lyso-PSs, partic-

elicits immunological responses (pro-inﬂammatory cytokine ularly (R)-2g and (R)-2h, produced highest secretion of TNF-a

secretion) in PPMs toward clearing this infection via a TLR4- and IL-6 from WT PPMs, and this pro-inﬂammatory cytokine

dependent pathway (Raetz and Whitﬁeld, 2002). In these assays, secretion was almost absent in TLR2-null PPMs (Figure 3B).

we found that the VLC (R)-Me-lyso-PS treatments produced

ABHD12 knockout mice display increased cerebellar microglio-

ꢂ60% pro-inﬂammatory cytokine secretion relative to similar sis, whereby the accumulation of VLC lyso-PSs speculatively

LPS treatments in PPMs, suggesting that like LPS, VLC (R)- causes this neuroinﬂammatory phenotype (Blankman et al.,

Me-lyso-PSs are potent immunological activators in mammalian 2013). To test whether VLC lyso-PSs indeed causes

Cell Chemical Biology 28, 1–11, June 17, 2021

5

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Figure 4. LC lyso-PSs activate macrophages

through a putative GPCR

(A–C) (A) Intracellular cAMP, (B) relative cytosolic

Ca²⁺levels, and (C) percentage of phosphorylated

ERK from WT PPMs following treatment with vehicle

(DMSO) or LPS or (R)-2a–2h (1 mM, 10 min, 37^ꢀC). All

data are presented as mean SD (n = 4–5/group),

where MC denotes medium chain, LC long chain,

and VLC very long chain.

(D) Representative western blots on lysates from

brains (6-month-old mice) and LPS-treated (1 mM, 4

h, 37^ꢀC) PPMs harvested from WT (+/+) or ABHD12-

null (ꢁ/ꢁ) mice, showing enhanced phosphorylation

of ERK in both these tissues for the ABHD12-null

(ꢁ/ꢁ) genotype.

(E) Schematic representation summarizing the lyso-

PS signaling pathways by possibly two types of re-

ceptors on mammalian macrophages.

(cAMP) (Sugita et al., 2013), cytosolic cal-

cium (Ca²⁺) inﬂux (Park et al., 2005), and

heightened phosphorylation of the nodal

extracellular signal-regulated kinase (ERK)

(Lee et al., 2008; Sugita et al., 2013), and

we wanted to test whether Me-lyso-PSs

also produce these phenotypes in

mammalian macrophages. We found in

WT PPMs that cellular cAMP (Figure 4A),

relative cytosolic Ca²⁺levels (Figure 4B),

and ERK phosphorylation (Figure 4C)

neuroinﬂammation and if they signal through TLR2, we intrave- increased most profoundly upon treatment with LC (R)-Me-

nously injected a VLC (R)-Me-lyso-PS ((R)-2h, C24:0) into WT or lyso-PSs, especially (R)-2e (C18:0), and that medium-chain or

TLR2-null mice, and quantiﬁed the extent of neuroinﬂammation VLC (R)-Me-lyso-PS had negligible effects on these phenotypes.

by counting the number of activated microglia using an Surprisingly, the genetic deletion of TLR2 in PPMs (Figures S5A,

established immunohistochemical protocol (Singh et al., 2020; S5B, and S6C) or its pharmacological antagonism in PPMs (Fig-

Blankman et al., 2013). We found that systemic administration of ures S5C, S5D, and S6B) and human THP-1 macrophages (Fig-

(R)-2h (C24:0) robustly induced cerebellar microgliosis in WT but ures S5E, S5F, and S6B) showed no change on any of these phe-

not TLR2 knockout mice, suggesting that VLC (R)-Me-lyso-PSs notypes following (R)-Me-lyso-PS treatment. We also found that

signals through TLR2, causing neuroinﬂammation (Figures 3C the LC (R)-Me-lyso-PSs (R)-2e (C18:0) and the corresponding

and S4). As a positive control, we used LPS (which signals through canonical lyso-PS 1b (C18:0) behaved identically in all these as-

TLR4) (Raetz and Whitﬁeld, 2002) and intravenously injected it at says (Figures S6D–S6F). Not surprisingly, upon treating PPMs

the same dose and time, and found that the extent of neuroinﬂam- with (S)-Me-lyso-PSs (Figures S6A, S6E, and S6F), the corre-

mation in WT and TLR2-null mice was almost identical (Figures 3C sponding free fatty acids or other lysophospholipids with lipid

and S4). Furthermore, we found that relative to the positive control tail (C18:0) (Figures S6D–S6F), failed to elicit any signiﬁcant

LPS, the VLC (R)-Me-lyso-PS (R)-2h (C24:0) produced compara- response in any of these assays under similar treatment condi-

ble neuroinﬂammation (Figures 3C and S4). Interestingly, a LC tions, suggesting that these phenotypes in mammalian macro-

(R)-Me-lyso-PS ((R)-2e, C18:0) intravenous injection for the same phages are speciﬁc to LC (R)-Me-lyso-PS (particularly C18:0).

dosing regimen did not produce any neuroinﬂammation in WT In all these assays, we found that the pharmacological Me-

mice (Figures S4A and S4B). Taken together, our results provide lyso-PS treatments did not affect cell viability.

compelling in vivo evidence that VLC lyso-PSs (but not LC lyso-

Of note, brains and LPS-treated PPMs derived from ABHD12-

PSs) signalthrough TLR2, elicita robustpro-inﬂammatory immune null mice (where, in absolute concentrations, 1b [C18:0 lyso-PS]

response, and are likely responsible for the neuroinﬂammation is the most abundant, and signiﬁcantly deregulated lyso-PS

observed in ageing ABHD12 knockout mice, the murine model of [Blankman et al., 2013; Kamat et al., 2015]) have markedly

PHARC (Blankman et al., 2013).

more ERK phosphorylation compared with WT control (Fig-

ure 4D). Given the heightened intracellular cAMP, cytosolic

Ca²⁺inﬂux, phosphorylation of ERK, and inability of pharmaco-

logical antagonism or genetic disruption of TLR2 to affect any

Mammalian macrophages have a cryptic lyso-PS

receptor

Mammalian macrophages in response to lyso-PS produce of these phenotypes, strongly supports the existence of another

increased intracellular cyclic adenosine 5⁰-monophosphate cryptic lyso-PS receptor on mammalian macrophages (in

6

Cell Chemical Biology 28, 1–11, June 17, 2021

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Figure 5. LC lyso-PSs robustly cause hista-

mine release from primary mast cells

(A) Representative microscopy image showing

degranulation of PCMCs after 1a (C16:0 lyso-PS) or

(R)-2d (C16:0) treatment (both 1 mM, 30 min, 37^ꢀC) in

the presence of concanavalin A (Con-A). Scale bars,

100 mm.

(B) Histamine release proﬁle from PCMCs treated

with 1a (C16:0 lyso-PS), (R)-2d (C16:0), or (S)-2d

(C16:0) (1 nM to 10 mM, 30 min, 37^ꢀC, n = 3/data

point).

(C) Dose-response and EC₅₀values for histamine

release from PCMCs following treatment with the

different (R)-Me-lyso-PSs (1 nM to 10 mM, 30 min,

37^ꢀC, n = 3/data point).

ceptor on PCMCs (likely GPR34 [Iwashita

et al., 2009; Sugo et al., 2006; Ikubo et al.,

2015], Figure S3A) is stereoselective in

lyso-PS recognition. Next, we performed

exhaustive dose-response studies with

the (R)-Me-lyso-PS library and found that

(R)-2d (C16:0, EC₅₀= 179 20 nM) most

potently induced histamine release during

PCMC degranulation, with other LC (R)-

Me-lyso-PSs (R)-2e (C18:0, EC₅₀= 311

37 nM) and (R)-2c (C14:0, EC₅₀= 645

addition to TLR2) (Figure 4E). We speculate, based on aforemen- 75 nM) following suit (Figure 5C). This SAR study showed that

tioned phenotypes, that this as yet unknown receptor is likely an medium-chain (<C12) or VLC (RC20) (R)-Me-lyso-PSs failed to

unannotated GPCR (as putative lyso-PS GPCRs [Inoue et al., induce signiﬁcant PCMC degranulation (EC₅₀> 2.5 mM) (Fig-

2012] are absent in macrophages [Figures S3A and S3B]) that ure 5C), suggesting that besides the headgroup, the recognition

prefers LC lyso-PSs (particularly C18:0) as ligands to produce of lipid tail is another major factor contributing to histamine

its downstream biological effects (Figure 4E).

release from mast cells during degranulation. In this SAR study,

we found that the pharmacological Me-lyso-PS treatments had

no effect on cell viability.

LC lyso-PSs robustly induce mast cell degranulation

The release of histamine during mast cell degranulation is an

extensively investigated lyso-PS-mediated immunological ABHD12 is the major lyso-PS lipase in primary mast cells

response (Sugo et al., 2006; Iwashita et al., 2009; Shanbhag ABHD12 is a major lyso-PS lipase in different immune cells (Oga-

et al., 2020), and yet contribution of individual lyso-PSs to this sawara et al., 2018; Kamat et al., 2015), but its biochemical func-

phenotype remain poorly understood. To address this, we tion in primary mast cells remains unknown. Western blot analysis

generated peritoneal-derived cultured mast cells (PCMCs) (Figure 6A) and diminished lyso-PS lipase activity (Figure 6B)

(Meurer et al., 2016), and conﬁrmed their ability to degranulate conﬁrmed the loss of ABHD12 in PCMCs derived from

in the presence of 1a (C16:0 lyso-PS) or the corresponding (R)- ABHD12-null mice. We also found that ABHD12-null PCMCs

Me-lyso-PS, (R)-2d (C16:0) in the presence of concanavalin A secreted signiﬁcantly more lyso-PS (ꢂ2-fold more 1a [C16:0]

(Con-A) (Figure 5A) (Iwashita et al., 2009; Sugo et al., 2006). and 1b [C18:0]) compared with WT PCMCs (Figure 6C). Given

Con-A is a well-characterized metalloprotein lectin of interest the diminished lyso-PS lipase activity, we postulated that 1a

to immunologists, as it speciﬁcally binds mannosyl and/or gluco- (C16:0 lyso-PS) treatment would cause ABHD12-null PCMCs to

syl units of surface receptors of immune cells, and in doing so degranulate at a lower dose of 1a, and thereby release histamine

stabilizes receptor conformations that, in the context of PCMCs, more efﬁciently than WT PCMCs. Indeed, ABHD12-null PCMCs

facilitate lyso-PS-dependent histamine release (Lawson et al., (EC₅₀= 47 17 nM) released histamine more effectively than

1978; Sullivan et al., 1975). Having shown their ability to degranu- WT PCMCs (EC₅₀= 135 32 nM) upon similar 1a (C16:0 lyso-

late, we quantitatively measured the histamine release from PS) treatment (Figure 6D). Finally, we measured serum histamine

PCMCs using an established LC-MS method (Chimalakonda concentrations following intravenous (R)-2d (C16:0) injection in

et al., 2015) (Figure S7). We found that the half-maximal effective WT or ABHD12 knockout mice and found that ABHD12-null

concentration (EC₅₀) toward inducing histamine release during mice displayed heightened circulating histamine concentrations

degranulation in PCMCs for 1a (C16:0 lyso-PS, EC₅₀= 140

(ꢂ3-fold) compared with WT mice (Figure 6E). These results

22 nM) and (R)-2d (C16:0, EC₅₀= 179 20 nM) were comparable, together conﬁrm ABHD12’s role as a major lyso-PS lipase in pri-

while the unnatural (S)-2d (C16:0, EC₅₀> 10 mM) failed to pro- mary mast cells, where by regulating serum lyso-PS levels (Fig-

duce this phenotype (Figure 5B), suggesting that the putative re- ure 6E) it controls systemic histamine release (Figure 6E).

Cell Chemical Biology 28, 1–11, June 17, 2021

7

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Figure 6. ABHD12 controls concentrations of

LC lyso-PSs in primary mast cells, and

thereby serum histamine levels

(A and B) (A) Representative western blot and (B)

lyso-PS lipase activity assay (100 mM 1a [C16:0 lyso-

PS], 30 min, 37^ꢀC, n = 3/group), showing the loss of

ABHD12 in PCMCs derived from ABHD12 knockout

(ꢁ/ꢁ) mice.

(C) Concentrations of lyso-PS secreted from

cultured PCMCs derived from WT (+/+) or ABHD12

knockout (ꢁ/ꢁ) mice, showing signiﬁcantly

increased 1a (C16:0 lyso-PS) and 1b (C18:0 lyso-

PS) secretion from ABHD12-null PCMCs (n = 4/

group).

(D) Histamine release proﬁles from PCMCs derived

from WT (+/+) or ABHD12 knockout (ꢁ/ꢁ) mice

treated with 1a (C16:0 lyso-PS) (1 nM to 10 mM,

30 min, 37^ꢀC, n = 3/data point).

(E) Left: serum histamine levels in (+/+) or (ꢁ/ꢁ) mice

following intravenous injections with vehicle (PBS) or

(R)-2d (C16:0) (1 mg/kg, 2 h), showing increased

serum histamine concentrations in the (ꢁ/ꢁ) mice

compared with (+/+) controls for both treatments

(n = 5/group). Interestingly, systemic administration

of (R)-2d (C16:0) in (+/+) mice produces signiﬁcantly

more circulating serum histamine compared with the

vehicle group, showing that (R)-2d (C16:0) by itself

can induce histamine release from mast cells in

in vivo settings. Right: serum lyso-PS concentra-

tions in (+/+) or (ꢁ/ꢁ) mice, showing increased concentrations of circulating 1a (C16:0 lyso-PS) and 1b (C18:0 lyso-PS) in (ꢁ/ꢁ) mice, corroborating the increased

serum histamine levels seen in these mice (n = 5/group). All data are presented as mean SD from at least n independent experiments. **p < 0.01, ***p < 0.001

versus (+/+) group by Student’s two-tailed unpaired parametric t test.

DISCUSSION

as excellent bioequivalent surrogates for the canonical (R)-

lyso-PSs, while the corresponding (S)-Me-lyso-PSs are biologi-

Signaling lipids are potent hormone-like biological molecules cally inactive. Since VLC lyso-PSs have never been tested

that regulate several important physiological processes in mam- against ABHD12, despite their accumulation in the brains of

mals, and their deregulation often has detrimental conse- ABHD12-null mice (Blankman et al., 2013; Singh et al., 2020),

quences that eventually manifest as disease in humans. Given here, by assaying the Me-lyso-PS library, we show conclusively

their direct link to human diseases, the lyso-PSs have been that the ABHD12-catalyzed lyso-PS lipase reaction is stereose-

recently emerged as yet another important class of signaling lective and prefers VLC lyso-PSs as substrates (Figure 1).

lipids (Shanbhag et al., 2020). However, unlike the established Furthermore, we show through biochemical assays that the

signaling lipids (e.g., 2-AG, AEA, S1P), in vivo lyso-PSs exist (R)-Me-lyso-PSs are acted upon by cellular carboxylesterases

esteriﬁed with different lipid tails ranging from medium-chain and converted to the corresponding canonical (R)-lyso-PSs,

(C10–C14) to LC (C16–C20) to VLC (>C20) fatty acids (Kamat and that therefore, biologically, the (R)-Me-lyso-PSs function

et al., 2015; Blankman et al., 2013; Singh et al., 2020). Given as prodrug-like surrogates for lyso-PSs in primary macrophages

the vast diversity in their in vivo content, it remains unclear as and mast cells (Figure 2).

to which of these lyso-PSs have signaling functions and how

Next, we wanted to understand the distinct role played by the

they inﬂuence, regulate, and/or modulate different immunolog- lipid tail of individual lyso-PSs in the activation of macrophages

ical processes. The lack of detailed SAR studies for these immu- and mast cell degranulation, as these immunological processes

nological phenotypes for lyso-PSs stems largely from the limited have been extensively studied and reported to be regulated by

commercial availability and reported synthetic strategies toward lyso-PSs (Shanbhag et al., 2020). In mammalian macrophages,

making them, especially the medium-chain and VLC variants.

we ﬁnd that VLC Me-lyso-PSs produce the highest secretion of

In this paper, we report the synthesis of Me-lyso-PSs having pro-inﬂammatory cytokines and that these pro-inﬂammatory re-

(R)- or (S)-stereochemistry at the sn-2 position of the glycerol sponses are higher in ABHD12-null macrophages (Figure 2),

backbone with varying lipid tails ranging from medium-chain to given their diminished lyso-PS lipase activity (Kamat et al.,

LC to VLC (Figure 1 and Data S1). Since the synthesis of lyso- 2015). Next, using pharmacological tools and genetic models,

PSs has proved challenging, and that of VLC lyso-PSs has not we show conclusively that VLC Me-lyso-PSs (but not LC lyso-

been reported to the best of our knowledge, our synthetic strat- PSs) orchestrate robust pro-inﬂammatory responses via a

egy allows the performance of rigorous SAR studies in diverse TLR2-dependent pathway to cause neuroinﬂammation (Fig-

assays to delineate the speciﬁc function of the lipid tail in modu- ure 3). That VLC lyso-PSs signal through TLR2 and cause neuro-

lating a phenotype regulated by lyso-PSs. In all our biochemical inﬂammation (Figures 3 and S4) raises an intriguing possibility,

and immunological assays, we ﬁnd that (R)-Me-lyso-PSs serve corroborating a recent report (Ogasawara et al., 2018) that the

8

Cell Chemical Biology 28, 1–11, June 17, 2021

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

human neurological disease PHARC may itself be an autoim- different physiological processes. Here, we report a syn-

mune disease. Furthermore, we ﬁnd that mammalian macro- thetic strategy toward making methyl esters of lyso-PSs

phages produce increased intracellular cAMP, cytosolic Ca²⁺(Me-lyso-PSs) with varying lipid tails, which serve as excel-

ﬂux, and heightened ERK phosphorylation in response to LC lent prodrug-like biological surrogates for lyso-PSs, and

lyso-PS treatments (particularly C18:0 lyso-PS) (Figure 4) via a focus on elucidating the role that the lipid tail of this lyso-

TLR2-independent pathway (Figures S5 and S6), and we specu- phospholipid plays in regulating various immunological

late that there is another cryptic receptor, likely a GPCR, that processes. Speciﬁcally, we study the lyso-PS-mediated

prefers LC lyso-PSs (particularly C18:0 lyso-PS) as ligands to activation of macrophages and mast cell degranulation pro-

produce these biological activities (Figure 4). We also show cesses, and show through detailed SAR studies that the lipid

that LC lyso-PSs (especially C16:0 lyso-PS), but not medium- tail has profound effects on the phenotypical outputs of

chain or VLC lyso-PSs, efﬁciently cause the release of histamine these important immunological processes. Our ﬁndings

from mast cells during degranulation, suggesting that the puta- thus illuminate a physiological balance between long-chain

tive lyso-PS receptor on mast cells (possibly GPR34 [Sugo and very-long-chain lyso-PSs, intricately regulated by the

et al., 2006]) prefers LC lyso-PSs (particularly C16:0 lyso-PS) lyso-PS lipase ABHD12, and that disrupting this ﬁne-tuned

as ligands to drive this immunological process (Figure 5). Finally, homeostasis results in immunological outputs that have

we report that ABHD12 is the major lyso-PS lipase in primary detrimental pathological consequences in humans.

mast cells, where it controls the secreted lyso-PS levels, and

its deletion in mice results in elevated levels of serum histamine STAR+METHODS

and lyso-PS (Figure 6).

Projecting ahead, we propose that the functional antagonism Detailed methods are provided in the online version of this paper

of TLR2 might provide an excellent therapeutic paradigm in and include the following:

treating PHARC. This premise can be tested genetically by

d KEY RESOURCES TABLE

generating and characterizing ABHD12-TLR2 dual-knockout

mice, and pharmacologically by discovering much needed in vivo

d RESOURCE AVAILABILITY

B Lead contact

active TLR2 functional antagonists. In addition, the annotation of

ABHD12 as a major lyso-PS lipase in primary mast cells provides

B Materials availability

B Data and code availability

yet another avenue in understanding the immunomodulatory

properties and the functional crosstalk of lyso-PSs between

d EXPERIMENTAL MODEL AND SUBJECT DETAILS

B Mice

different immune cells particularly in the context of allergies

and autoimmune conditions (Kelkar et al., 2019; Ogasawara

B Mammalian cell lines

d METHOD DETAILS

et al., 2018). Finally, we would like to note that a major short-

coming of our study is the inability of our synthetic strategy in

B Reagents

B Lyso-PS lipase substrate assays

making Me-lyso-PSs with unsaturated fatty acid chains. There-

fore, the development of such a synthetic methodology would

B Mammalian cells lipid treatments

B Western blot analysis

enable making Me-lyso-PSs with unsaturated fatty acid chains

and could also be leveraged to generate lyso-PS probes with

B ELISA assays

B RT-PCR

suitable recently reported biorthogonal handles (Niphakis et al.,

2015). Such bifunctional lyso-PS probes in tandem with

B Immunohistochemical analysis

B Histamine estimation from mast cells

advanced MS-based chemoproteomics platforms (Niphakis

et al., 2015; Parker et al., 2017) will greatly facilitate the identiﬁ-

B Lyso-PS measurements

B Synthesis and compound characterization

cation of hitherto unknown lyso-PS protein ligands and/or recep-

tors, and this emerging knowledge will certainly expand our bio-

d QUANTIFICATION AND STATISTICAL ANALYSIS

logical understanding of this underexplored immunomodulatory

lipid class.

SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j.

SIGNIFICANCE

chembiol.2021.01.008.

ACKNOWLEDGMENTS

Lysophospholipids are potent hormone-like biological me-

diators that regulate many important physiological pro-

cesses in mammals. Recently, the lysophosphatidylserines

(lyso-PSs) have emerged as yet another class of signaling ly-

sophospholipids, with potent bioactivities in the central ner-

vous and immune system in mammals, and deregulation in

their metabolism has been linked to neurological and auto-

immune disorders in humans. However, challenges in mak-

ing lyso-PSs synthetically, and limited commercial sources

for this lipid class, have greatly hampered any exhaustive

structure-activity relationship (SAR) studies toward mecha-

nistically understanding the role that lyso-PSs play in

This work was supported by a DBT/Wellcome Trust India Alliance Fellowship

(grant number IA/I/15/2/502058) awarded to S.S.K., a Department of Science

and Technology (DST) Fund for Improvement of S&T Infrastructure (grant num-

ber SR/FST/LSII-043/2016) to the IISER Pune Biology Department, and a J.C.

Bose National Fellowship from the Science & Engineering Research Board

(SERB) (grant number SB/S2/JCB-025/2016) awarded to K.N.B. N.K. acknowl-

edges a SERB postdoctoral fellowship, and M.S. acknowledges a graduate stu-

dent fellowship from the Council of Scientiﬁc and Industrial Research. Benjamin

F. Cravatt of The Scripps Research Institute is thanked for providing the

ABHD12 knockout mice used in this study. The National Facility for Gene Func-

tion in Health and Disease (NFGFHD) at IISER Pune (supported by a grant from

the Department of Biotechnology, Govt. of India; BT/INF/22/SP17358/2016) and

Cell Chemical Biology 28, 1–11, June 17, 2021

9

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Central Animal Facility at IISc Bangalore are thanked for maintaining and

providing mice for this study. Sagar Tarate and Dhanashree Kelkar are thanked

for technical assistance. Vineeta Bal, Satyajit Rath, Girish Deshpande, Girish

Ratnaparkhi, and Nishad Matange are thanked for reading and providing critical

inputs to the manuscript.

Fowler, C.J., Holt, S., Nilsson, O., Jonsson, K.O., Tiger, G., and Jacobsson,

S.O. (2005). The endocannabinoid signaling system: pharmacological and

therapeutic aspects. Pharmacol. Biochem. Behav. 81 , 248–262.

Frasch, S.C., and Bratton, D.L. (2012). Emerging roles for lysophosphatidylser-

ine in resolution of inﬂammation. Prog. Lipid Res. 51 , 199–207.

Gardell, S.E., Dubin, A.E., and Chun, J. (2006). Emerging medicinal roles for ly-

AUTHOR CONTRIBUTIONS

sophospholipid signaling. Trends Mol. Med. 12 , 65–75.

N.K., S.S., T.S., and A.J. performed all the biochemical studies. M.S. and A.M.

synthesized and characterized all the compounds; A.M. and H.C. supervised

the synthesis. K.N.B. provided the TLR2 knockout mice for this study. S.S.K.

conceived the project, acquired funding, analyzed the data and wrote the pa-

per with input from all authors.

Gonzalez-Cabrera, P.J., Brown, S., Studer, S.M., and Rosen, H. (2014). S1P

signaling: new therapies and opportunities. F1000prime Rep. 6 , 109.

Holla, S., Prakhar, P., Singh, V., Karnam, A., Mukherjee, T., Mahadik, K.,

Parikh, P., Singh, A., Rajmani, R.S., Ramachandra, S.G., and Balaji, K.N.

(2016). MUSASHI-mediated expression of JMJD3, a H3K27me3 demethylase,

is involved in foamy macrophage generation during mycobacterial infection.

PLoS Pathog. 12 , e1005814.

DECLARATION OF INTERESTS

The authors declare no competing interests.

Hosono, H., Aoki, J., Nagai, Y., Bandoh, K., Ishida, M., Taguchi, R., Arai, H.,

and Inoue, K. (2001). Phosphatidylserine-speciﬁc phospholipase A1 stimu-

lates histamine release from rat peritoneal mast cells through production of

2-acyl-1-lysophosphatidylserine. J. Biol. Chem. 276 , 29664–29670.

Received: August 31, 2020

Revised: November 30, 2020

Accepted: January 6, 2021

Published: February 10, 2021

Ikubo, M., Inoue, A., Nakamura, S., Jung, S.J., Sayama, M., Otani, Y.,

Uwamizu, A., Suzuki, K., Kishi, T., Shuto, A., et al. (2015). Structure-activity re-

lationships of lysophosphatidylserine analogs as agonists of G-protein-

coupled receptors GPR34, P2Y10, and GPR174. J. Med. Chem. 58 ,

4204–4219.

REFERENCES

Barnes, M.J., Li, C.M., Xu, Y., An, J., Huang, Y., and Cyster, J.G. (2015). The

lysophosphatidylserine receptor GPR174 constrains regulatory T cell develop-

ment and function. J. Exp. Med. 212 , 1011–1020.

Inoue, A., Ishiguro, J., Kitamura, H., Arima, N., Okutani, M., Shuto, A.,

Higashiyama, S., Ohwada, T., Arai, H., Makide, K., and Aoki, J. (2012).

TGFalpha shedding assay: an accurate and versatile method for detecting

GPCR activation. Nat. Methods 9 , 1021–1029.

Benjamini, Y., Krieger, A.M., and Yekutieli, D. (2006). Adaptive linear step-up

procedures that control the false discovery rate. Biometrika 93 , 491–507.

Ishii, I., Fukushima, N., Ye, X., and Chun, J. (2004). Lysophospholipid recep-

Blankman, J.L., and Cravatt, B.F. (2013). Chemical probes of endocannabi-

tors: signaling and biology. Annu. Rev. Biochem. 73 , 321–354.

noid metabolism. Pharmacol. Rev. 65 , 849–871.

Iwashita, M., Makide, K., Nonomura, T., Misumi, Y., Otani, Y., Ishida, M.,

Taguchi, R., Tsujimoto, M., Aoki, J., Arai, H., and Ohwada, T. (2009).

Synthesis and evaluation of lysophosphatidylserine analogues as inducers of

mast cell degranulation. Potent activities of lysophosphatidylthreonine and

its 2-deoxy derivative. J. Med. Chem. 52 , 5837–5863.

Blankman, J.L., Long, J.Z., Trauger, S.A., Siuzdak, G., and Cravatt, B.F.

(2013). ABHD12 controls brain lysophosphatidylserine pathways that are de-

regulated in a murine model of the neurodegenerative disease PHARC.

Proc. Natl. Acad. Sci. U S A 110 , 1500–1505.

Blankman, J.L., Simon, G.M., and Cravatt, B.F. (2007). A comprehensive pro-

ﬁle of brain enzymes that hydrolyze the endocannabinoid 2-arachidonoylgly-

cerol. Chem. Biol. 14 , 1347–1356.

Joshi, A., Shaikh, M., Singh, S., Rajendran, A., Mhetre, A., and Kamat, S.S.

(2018). Biochemical characterization of the PHARC-associated serine hydro-

lase ABHD12 reveals its preference for very-long-chain lipids. J. Biol. Chem.

293 , 16953–16963.

Chimalakonda, K.C., Pang, E., Weaver, J.L., Howard, K.E., Patel, V., and

Boyne, M.T. (2015). Development and validation of a liquid-chromatography

tandem mass spectrometry method to determine in vitro and in vivo histamine

release. J. Pharm. Biomed. Anal. 102 , 494–499.

Kamat, S.S., Camara, K., Parsons, W.H., Chen, D.H., Dix, M.M., Bird, T.D.,

Howell, A.R., and Cravatt, B.F. (2015). Immunomodulatory lysophosphatidyl-

serines are regulated by ABHD16A and ABHD12 interplay. Nat. Chem. Biol.

11 , 164–171.

Chu, X., Shen, M., Xie, F., Miao, X.J., Shou, W.H., Liu, L., Yang, P.P., Bai, Y.N.,

Zhang, K.Y., Yang, L., et al. (2013). An X chromosome-wide association anal-

ysis identiﬁes variants in GPR174 as a risk factor for Graves’ disease. J. Med.

Genet. 50 , 479–485.

Kelkar, D.S., Ravikumar, G., Mehendale, N., Singh, S., Joshi, A., Sharma, A.K.,

Mhetre, A., Rajendran, A., Chakrapani, H., and Kamat, S.S. (2019). A chemical-

genetic screen identiﬁes ABHD12 as an oxidized-phosphatidylserine lipase.

Nat. Chem. Biol. 15 , 169–178.

Contos, J.J.A., Ishii, I., and Chun, J. (2000). Lysophosphatidic acid receptors.

Mol. Pharm. 58 , 1188–1196.

Kimura, I., Moritani, Y., and Tanizaki, Y. (1973). Basophils in bronchial asthma

Dennis, E.A. (2016). Lipidomics in disease and drug discovery. Faseb J. 30,

with reference to reagin-type allergy. Clin. Allergy 3 , 195–202.

https://doi.org/10.1096/fasebj.30.1_supplement.114.3.

Lawson, D., Fewtrell, C., and Raff, M.C. (1978). Localized mast cell degranu-

lation induced by concanavalin A-Sepharose beads. Implications for the

Ca2+ hypothesis of stimulus-secretion coupling. J. Cell Biol. 79 , 394–400.

Dennis, E.A., and Norris, P.C. (2015). Eicosanoid storm in infection and inﬂam-

mation. Nat. Rev. Immunol. 15 , 511–523.

Fahy, E., Cotter, D., Sud, M., and Subramaniam, S. (2011). Lipid classiﬁcation,

structures and tools. Biochim. Biophys. Acta 1811 , 637–647.

Lee, S.Y., Lee, H.Y., Kim, S.D., Jo, S.H., Shim, J.W., Lee, H.J., Yun, J., and

Bae, Y.S. (2008). Lysophosphatidylserine stimulates chemotactic migration

in U87 human glioma cells. Biochem. Biophys. Res. Commun. 374 , 147–151.

Fahy, E., Subramaniam, S., Brown, H.A., Glass, C.K., Merrill, A.H., Jr., Murphy,

R.C., Raetz, C.R., Russell, D.W., Seyama, Y., Shaw, W., et al. (2005). A

comprehensive classiﬁcation system for lipids. J. Lipid Res. 46 , 839–861.

Liang, Q.L., Wu, Q.A., Jiang, J.H., Duan, J.A., Wang, C., Smith, M.D., Lu, H.,

Wang, Q., Nagarkatti, P., and Fan, D.P. (2011). Characterization of

Sparstolonin B, a Chinese herb-derived compound, as a selective toll-like re-

ceptor antagonist with potent anti-inﬂammatory properties. J. Biol. Chem. 286 ,

26470–26479.

Fiskerstrand, T., H’mida-Ben Brahim, D., Johansson, S., M’zahem, A.,

Haukanes, B.I., Drouot, N., Zimmermann, J., Cole, A.J., Vedeler, C.,

Bredrup, C., et al. (2010). Mutations in ABHD12 cause the neurodegenerative

disease PHARC: an inborn error of endocannabinoid metabolism. Am. J. Hum.

Genet. 87 , 410–417.

Liang, Q.L., Yu, F., Cui, X.D., Duan, J.A., Wu, Q.N., Nagarkatti, P., and Fan,

D.P. (2013). Sparstolonin B suppresses lipopolysaccharide-induced inﬂam-

mation in human umbilical vein endothelial cells. Arch. Pharmacal Res. 36 ,

890–896.

Fiskerstrand, T., Knappskog, P., Majewski, J., Wanders, R.J., Boman, H., and

Bindoff, L.A. (2009). A novel Refsum-like disorder that maps to chromosome

20. Neurology 72 , 20–27.

10 Cell Chemical Biology 28, 1–11, June 17, 2021

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Lloret, S., and Moreno, J.J. (1995). Ca²⁺inﬂux, phosphoinositide hydrolysis,

and histamine release induced by lysophosphatidylserine in mast cells.

J. Cell Physiol. 165 , 89–95.

Schindelin, J., Rueden, C.T., Hiner, M.C., and Eliceiri, K.W. (2015). The ImageJ

ecosystem: an open platform for biomedical image analysis. Mol. Reprod.

Dev. 82 , 518–529.

Mallik, S., Prasad, R., Bhattacharya, A., and Sen, P. (2018). Synthesis of phos-

phatidylserine and its stereoisomers: their role in activation of blood coagula-

tion. ACS Med. Chem. Lett. 9 , 434–439.

Shanbhag, K., Mhetre, A., Khandelwal, N., and Kamat, S.S. (2020). The lyso-

phosphatidylserines—an emerging class of signalling lysophospholipids.

J. Membr. Biol. 253 , 381–397.

Meurer, S.K., Ness, M., Weiskirchen, S., Kim, P., Tag, C.G., Kauffmann, M.,

Huber, M., and Weiskirchen, R. (2016). Isolation of mature (peritoneum-

derived) mast cells and immature (bone marrow-derived) mast cell precursors

from mice. PLoS One 11 , e0158104.

Singh, S., Joshi, A., and Kamat, S.S. (2020). Mapping the neuroanatomy

of ABHD16A, ABHD12, and lysophosphatidylserines provides new insights

into the pathophysiology of the human neurological disorder PHARC.

Biochemistry 59 , 2299–2311.

Napier, C., Mitchell, A.L., Gan, E., Wilson, I., and Pearce, S.H.S. (2015). Role of

the X-linked gene GPR174 in autoimmune Addison’s disease. J. Clin.

Endocrinol. Metab. 100 , E187–E190.

Sugita, K., Yamamura, C., Tabata, K., and Fujita, N. (2013). Expression of

orphan G-protein coupled receptor GPR174 in CHO cells induced morpholog-

ical changes and proliferation delay via increasing intracellular cAMP.

Biochem. Biophys. Res. Commun. 430 , 190–195.

Navia-Paldanius, D., Savinainen, J.R., and Laitinen, J.T. (2012). Biochemical

and pharmacological characterization of human alpha/beta-hydrolase domain

containing 6 (ABHD6) and 12 (ABHD12). J. Lipid Res. 53 , 2413–2424.

Sugo, T., Tachimoto, H., Chikatsu, T., Murakami, Y., Kikukawa, Y., Sato, S.,

Kikuchi, K., Nagi, T., Harada, M., Ogi, K., et al. (2006). Identiﬁcation of a lyso-

phosphatidylserine receptor on mast cells. Biochem. Biophys. Res. Commun.

341 , 1078–1087.

Niphakis, M.J., Lum, K.M., Cognetta, A.B., 3rd, Correia, B.E., Ichu, T.A.,

Olucha, J., Brown, S.J., Kundu, S., Piscitelli, F., Rosen, H., and Cravatt, B.F.

(2015). A global map of lipid-binding proteins and their ligandability in cells.

Cell 161 , 1668–1680.

Sullivan, T.J., Greene, W.C., and Parker, C.W. (1975). Concanavalin A-

induced histamine release from normal rat mast cells. J. Immunol. 115 ,

278–282.

Ogasawara, D., Ichu, T.A., Vartabedian, V.F., Benthuysen, J., Jing, H., Reed,

A., Ulanovskaya, O.A., Hulce, J.J., Roberts, A., Brown, S., et al. (2018).

Selective blockade of the lyso-PS lipase ABHD12 stimulates immune re-

sponses in vivo. Nat. Chem. Biol. 14 , 1099–1108.

Szymanski, K., Miskiewicz, P., Pirko, K., Jurecka-Lubieniecka, B., Kula, D.,

Hasse-Lazar, K., Krajewski, P., Bednarczuk, T., and Ploski, R. (2014).

rs3827440,

a nonsynonymous single nucleotide polymorphism within

Park, K.S., Lee, H.Y., Kim, M.K., Shin, E.H., and Bae, Y.S. (2005).

Lysophosphatidylserine stimulates leukemic cells but not normal leukocytes.

Biochem. Biophys. Res. Commun. 333 , 353–358.

GPR174 gene in X chromosome, is associated with Graves’ disease in

Polish Caucasian population. Tissue Antigens 83 , 41–44.

Parker, C.G., Galmozzi, A., Wang, Y., Correia, B.E., Sasaki, K., Joslyn, C.M.,

Kim, A.S., Cavallaro, C.L., Lawrence, R.M., Johnson, S.R., et al. (2017).

Ligand and target discovery by fragment-based screening in human cells.

Cell 168 , 527–541 e29.

Tracey, T.J., Steyn, F.J., Wolvetang, E.J., and Ngo, S.T. (2018). Neuronal lipid

metabolism: multiple pathways driving functional outcomes in health and dis-

ease. Front. Mol. Neurosci. 11 , 10.

Van Der Kleij, D., Latz, E., Brouwers, J.F., Kruize, Y.C., Schmitz, M., Kurt-

Jones, E.A., Espevik, T., De Jong, E.C., Kapsenberg, M.L., Golenbock, D.T.,

et al. (2002). A novel host-parasite lipid cross-talk. Schistosomal lyso-phos-

phatidylserine activates toll-like receptor 2 and affects immune polarization.

J. Biol. Chem. 277 , 48122–48129.

Pathak, D., Mehendale, N., Singh, S., Mallik, R., and Kamat, S.S. (2018).

Lipidomics suggests a new role for ceramide synthase in phagocytosis. ACS

Chem. Biol. 13 , 2280–2287.

Raetz, C.R., and Whitﬁeld, C. (2002). Lipopolysaccharide endotoxins. Annu.

Rev. Biochem. 71 , 635–700.

Vance, J.E. (2015). Phospholipid synthesis and transport in mammalian cells.

Rajendran, A., Vaidya, K., Mendoza, J., Bridwell-Rabb, J., and Kamat, S.S.

(2020). Functional annotation of ABHD14B, an orphan serine hydrolase

enzyme. Biochemistry 59 , 183–196.

Trafﬁc 16 , 1–18.

Wu, C., Jin, X., Tsueng, G., Afrasiabi, C., and Su, A.I. (2016). BioGPS: building

your own mash-up of gene annotations and expression proﬁles. Nucleic Acids

Res. 44 , D313–D316.

Rosen, H., Stevens, R.C., Hanson, M., Roberts, E., and Oldstone, M.B. (2013).

Sphingosine-1-phosphate and its receptors: structure, signaling, and inﬂu-

ence. Annu. Rev. Biochem. 82 , 637–662.

Wymann, M.P., and Schneiter, R. (2008). Lipid signalling in disease. Nat. Rev.

Mol. Cell Biol. 9 , 162–176.

Rueden, C.T., Schindelin, J., Hiner, M.C., Dezonia, B.E., Walter, A.E., Arena,

E.T., and Eliceiri, K.W. (2017). ImageJ2: ImageJ for the next generation of sci-

entiﬁc image data. BMC Bioinformatics 18 , 529.

Yanagida, K., and Valentine, W.J. (2020). Druggable lysophospholipid

signaling pathways. Adv. Exp. Med. Biol. 1274 , 137–176.

Cell Chemical Biology 28, 1–11, June 17, 2021 11

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

STAR+METHODS

KEY RESOURCES TABLE

REAGENT or RESOURCE

Antibodies

SOURCE

IDENTIFIER

Rabbit, anti-phospho-p44/42 MAPK (ERK1/2) (20G11)

Rabbit, anti-p44/42 MAPK (ERK1/2) (137F5)

Rabbit, anti-GAPDH (EPR16891)

Cell Signaling Technology

Abcam

Clone: monoclonal, catalog# 4376, RRID: AB_331772

Clone: monoclonal, catalog# 4695, RRID: AB_390779

Clone: monoclonal, catalog# ab181602, RRID:

AB_2630358

Rabbit, anti-ABHD12 (EPR13683-72) (produced

recombinantly, animal free)

Abcam

Clone: monoclonal, catalog# ab182011, RRID:

Not available

Rabbit, anti-Iba1 (EPR16588)

Rabbit, anti-beta actin

Rabbit, anti-TLR2

Abcam

Clone: monoclonal, catalog# ab178846, RRID:

AB_2636859

Cloud-Clone Corp.

Novus Biologicals

Thermo Fisher Scientiﬁc

Vector Laboratories

Clone: polyclonal, catalog# CAB340Hu01, RRID:

Not available

Clone: polyclonal, catalog# NB100-56720, RRID:

AB_838993

Goat, anti-rabbit IgG (H + L) secondary antibody,

HRP conjugated

Clone: polyclonal, catalog# 31,460, RRID:

AB_228341

Horse, anti-rabbit IgG (H + L) secondary antibody,

Biotinylated

Clone: polyclonal, catalog# BP-1100, RRID:

Not available

Critical commercial assays

Mouse TNF-alpha DuoSet

Mouse IL-6 DuoSet

R&D Systems

Abcam

Catalog# DY410

Catalog# DY406

Catalog# DY210

Catalog# DY206

Catalog# ab65355

Catalog# ab112128

Catalog# ab126445

Human TNF-alpha DuoSet

Human IL-6 DuoSet

cAMP assay kit (competitive ELISA)

Flou-8 calcium assay kit – medium removal

Abcam

ERK1 (pT202/pY204) + ERK2 (pT185/pY187) +

total ERK1/2 ELISA kit

Abcam

Chemicals, peptides and recombinant proteins

C16:0 Lyso-PS

Avanti Polar Lipids

Sigma-Aldrich (Merck)

Polysciences

Catalog# 858142

Catalog# 858144

Catalog# 858143

Catalog# 858141

Catalog# 857128

Catalog# 858124

Catalog# 855775

Catalog# H8896

Catalog# S4751

Catalog# T9032

Catalog# L3024

Catalog# I4144

Catalog# S9915

Catalog# SML1767

Catalog# 46,058

Catalog# H7125

Catalog# C0412

Catalog# 24,765

C18:0 Lyso-PS

C18:1 Lyso-PS

C17:1 Lyso-PS

C18:0 Lyso-PA

C18:0 Lyso-PG

C18:0 Lyso-PC

cis-10-Heptadecenoic acid (C17:1 FFA)

Stearic acid

Thioglycollate medium

Lipopolysaccharide

Mouse interleukin-3

Mouse stem cell factor

Sparstolonin B (SpaB)

Carboxylesterase from porcine liver

Histamine

Concanavalin A (Con-A)

Polyethylenimine hydrochloride (PEI MAXꢀ)

(Continued on next page)

e1 Cell Chemical Biology 28, 1–11.e1–e6, June 17, 2021

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

Continued

REAGENT or RESOURCE

Experimental models: organisms/strains

Mouse: C57BL/6J

SOURCE

IDENTIFIER

The Jackson Laboratory

Scripps Research

RRID: IMSR_JAX:000664

N/A

Mouse: ABHD12 null

Mouse: TLR2 null

The Jackson Laboratory

RRID: IMSR_JAX:004650

Experimental models: cell lines

HEK293T cells

ATCC

RRID: CVCL_0063

RRID: CVCL_0006

THP-1 cells

Oligonucleotides

Primers for RT-PCR, see Table S1

Software and algorithms

This paper

N/A

Prism 7 (version 7.0e)

ImageJ v1.52u

GraphPad

NIH

https://www.graphpad.com/scientiﬁc-software/prism/

https://imagej.nih.gov/ij/

SciexOS (version 1.4.0.18067)

Sciex

https://sciex.com/products/software/sciex-os-

software-x120805

Other

Lipidomics data (mass spec)

This paper

Available from lead contact upon reasonable request

RESOURCE AVAILABILITY

Lead contact

Further information and materials request should be directed to and will be fulﬁlled only by the Lead Contact, Siddhesh S. Kamat

(siddhesh@iiserpune.ac.in).

Materials availability

Any compounds, or reagents described/generated in this study, are available with the Lead Contact upon request with a completed

Materials Transfer Agreement.

Data and code availability

All the data that supports the ﬁndings of this study are available in the paper (and its associated Supplemental information) or are

available from the Lead Contact on reasonable request. No unique codes were generated in this study.

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Mice

All mouse studies and experiments described in this paper have received formal approval from the Indian Institute of Science Edu-

cation and Research, Pune – Institutional Animal Ethics Committee (IISER-P IAEC) (application nos: IISER_Pune IAEC/2016_02/01,

and IISER_Pune IAEC/2019_2/07) constituted as per the guidelines and norms provided by the Committee for the Purpose of Control

and Supervision of Experiments in Animals (CPCSEA), Government of India. All experimental mice were housed in the National Fa-

cility for Gene Function in Health and Disease (NFGFHD), IISER Pune, and were studied between 2 – 6 months of age. All mice had ab

libitum access to food and water. Thioglycollate elicited primary peritoneal macrophages (PPMs) (Singh et al., 2020; Kelkar et al.,

2019) and peritoneal-derived cultured mast cells (PCMCs) (Meurer et al., 2016) were generated from mice 10 – 12 weeks of age, har-

vested from mice and cultured using established protocols. For all mouse experiments, equal number of male and female mice were

used, and age/sex matched littermates were used as controls. All strains of mice used in this study as reported in the Key re-

sources table.

Mammalian cell lines

HEK293T (RRID: CVCL_0063, female) and THP-1 (RRID: CVCL_0006, male) cells were purchased from the ATCC, and cultured in

RPMI1640 medium supplemented with 10% (v/v) FBS and antibiotics (1% (v/v) penicillin-streptomycin (MP Biomedicals)) at 37^ꢀC

with 5% (v/v) CO₂. Cell staining with 4⁰,6-diamidino-2-phenylindole (DAPI) was routinely performed to ensure that all cell lines

were devoid of any mycoplasma contamination. All cell lines were grown to ꢂ80% conﬂuency, and all live cells were estimated

on a TC20 automated cell counter using trypan blue reagent as per manufacturer’s protocol (Bio-Rad).

Cell Chemical Biology 28, 1–11.e1–e6, June 17, 2021 e2

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

METHOD DETAILS

Reagents

Unless otherwise mentioned, all the following materials used in the experiments described in this paper were purchased from the

commercial source mentioned in the parenthesis: chemicals, buffers and reagents (Sigma-Aldrich), commercial lipids and lipid stan-

dards (Avanti Polar Lipids), primary and secondary antibodies (Abcam), and tissue culture media and Fetal Bovine Serum (FBS)

(HiMedia).

Lyso-PS lipase substrate assays

Wild type (WT) human ABHD12 (hABHD12) or the catalytically inactive active site mutant S246A hABHD12 were transiently trans-

fected in HEK293T cells using a polyethylenimine-based (PEI MAXꢀ) transfection methodology (Joshi et al., 2018; Kamat et al.,

2015) along with mock (empty plasmid) controls, and the membrane lysates from these were prepared as reported earlier (Kelkar

et al., 2019; Joshi et al., 2018). Brieﬂy, HEK293T cells overexpressing either WT ABHD12 or S246A ABHD12 or ‘‘mock’’ control (empty

plasmid) were harvested by scraping 48 hours post-transfection, washed (3x) with cold sterile Dulbecco’s phosphate buffer saline

(DPBS) without calcium and magnesium (pH 7.2) (HiMedia), re-suspended 1 mL DPBS, and lysed by sonication. The cell debris (pel-

let) was discarded after centrifugation at 200g for 5 min at 4^ꢀC, and the resulting lysate (supernatant, ꢂ 900 mL) was centrifuged at

100,000g for 45 min at 4^ꢀC. The supernatant from this step was also discarded, and the pellet (membrane proteome) was washed (3x)

with cold sterile DPBS, and re-suspended to uniformity by pipetting in 500 mL of cold sterile DPBS. The protein concentration was

estimated using BCA Protein assay kit (Pierce), and the expression of hABHD12 in this cell membrane lysate was conﬁrmed by west-

ern blot analysis (Joshi et al., 2018). The brain membrane lysates from wild type and ABHD12 knockout mice were also prepared

using a protocol reported earlier (Joshi et al., 2018; Singh et al., 2020). Brieﬂy, the mice were anaesthetized using isoﬂurane, and

euthanized by cervical dislocation. The mouse brains were harvested, and washed (3x) using DBPS. Following this, a half brain

was suspended in 500 mL of cold sterile DBPS, and homogenized by a tissue homogenizer (Bullet Blender 24, Next Advance) using

1 scoop of glass beads (0.5 mm diameter, Next Advance) for 3 min at 4^ꢀC at an instrument speed setting of 8. To this brain homog-

enate, cold sterile DPBS (500 mL) was added, mixed to uniformity by pipetting, centrifuged at 1000g for 5 min at 4^ꢀC to precipitate the

tissue debris. The supernatant lysate (ꢂ700 mL) from this step was collected, and centrifuged at 100,000g for 45 min at 4^ꢀC. The re-

sulting supernatant was discarded and the pellet (membrane proteome) was washed with cold sterile DPBS (3x), re-suspended in

1 mL of cold sterile DPBS by pipetting, and the ﬁnal protein concentration of the brain membrane lysate was estimated using the

BCA Protein assay kit (Pierce). The lyso-PS lipase assays described in this paper were performed using an established LC-MS

method that measures the release of free fatty acids from lyso-PS substrates by the lipase activity of ABHD12 relative to an unnatural

fatty acid internal standard (C17:1 FFA, 0.5 nmol/sample) (Joshi et al., 2018; Singh et al., 2020). All assays done using 10 mg of

HEK293T membrane lysates transfected with hABHD12 or 20 mg of brain membrane lysates for 30 mins, as previous studies

have shown that the lyso-PS lipase assay is linear for this protein concentration and time over a range of substrate concentrations

(Joshi et al., 2018). All single concentration lipase (substrate) assays were done using 100 mM of the lipid substrate, while the enzyme

kinetic studies were done over the concentration range of 0 – 400 mM for the particular lipid that was assayed.

Mammalian cells lipid treatments

Primary peritoneal macrophages (PPMs) were freshly obtained from mice using a protocol previously reported by us (Singh et al.,

2020; Kelkar et al., 2019) with typical yields of (ꢂ10 x 10⁶) PPMs per mouse. Brieﬂy, PPMs were elicited by intraperitoneal injection

of aged thioglycollate broth (3% w/v) and 60 hours post-injection, the PPMs were harvested by a peritoneal lavage using 10 mL cold

sterile DBPS. Post-harvest, the cells were pelleted by centrifugation at 1400g for 5 mins, and re-suspended in 5 mL of RBC lysis

buffer for 10 mins at 4^ꢀC. After removal of the RBC lysis buffer by centrifugation at 1400g for 5 mins at 4^ꢀC, the PPMs were re-sus-

pended in 5 mL of RPMI 1640 medium supplemented with 10% (v/v) FBS and antibiotics (1% (v/v) penicillin-streptomycin (MP Bio-

medicals)) at 37^ꢀC with 5% (v/v) CO₂, following which, live PPMs were estimated on a TC20 automated cell counter using trypan blue

reagent as per manufacturer’s protocol (Bio-Rad) and plated at the required cell density as per the desired assay condition. PPMs

were allowed to adhere to the plates for 2 h at 37^ꢀC and 5% (v/v) CO₂, following which, the media was replaced to that required for the

desired assay. Equal number of male and female mice were used for the generation of PPMs. For the pro-inﬂammatory cytokine

measurements, macrophages (1 x 10⁶PPMs or 0.5 x 10⁶THP-1 cells) were plated in 6-well plates (Eppendorf) in 2 mL of

RPMI1640 media without any supplementation, and treated with vehicle (DMSO) or lipids [commercial lyso-PSs or (R)-Me-lyso-

PSs or (S)-Me-lyso-PSs or free fatty acids, or other C18:0 lysophospholipids (lyso-PA, lyso-PG and lyso-PC)] or LPS (positive assay

control) all at 1 mM ﬁnal concentration for 4 hours at 37^ꢀC. Following this treatment, the media (1.5 mL) was collected and stored at –

80^ꢀC till the analysis was completed. The pro-inﬂammatory cytokines (TNF-a and IL-6) were quantitatively measured using standard

single analyte ELISA assays as per manufacturer’s instructions (R&D Systems). For the cellular cyclic adenosine 5⁰-monophosphate

(cAMP), cytosolic calcium (Ca²⁺) and ERK phosphorylation measurements, macrophages (4 x 10⁶PPMs or 2 x 10⁶THP-1 cells) were

plated in 6-well plates (Eppendorf) in 3 mL of RPMI1640 media without any supplementation, and treated with vehicle (DMSO) or

lipids [commercial lyso-PSs or (R)-Me-lyso-PSs or (S)-Me-lyso-PSs or free fatty acids, or other C18:0 lysophospholipids (lyso-PA,

lyso-PG and lyso-PC)] or LPS (positive assay control) all at 1 mM ﬁnal concentration for 10 mins at 37^ꢀC. Following this treatment,

the assay plates were centrifuged at 1400g for 5 min to pellet the live cells, and separate the media. The media was removed by pi-

petting, following which, the cells were washed with sterile cold DPBS (x 3 times) at 4^ꢀC, and lysed by sonication in 500 mL cold sterile

e3 Cell Chemical Biology 28, 1–11.e1–e6, June 17, 2021

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

DPBS at 4^ꢀC. The cell lysates were deproteinized by passing them through a 3000 (3-kDa) molecular weight ﬁlter, and the ﬂow

through was eventually used to estimate the cellular levels of cAMP and free cytosolic Ca²⁺. For phosphorylated ERK measurements,

total cell lysates without the deproteinization step were used. The changes (increase) in cellular cAMP, free cytosolic Ca²⁺and ERK

phosphorylation in response to the aforementioned treatments were quantitatively estimated using standard competitive ELISA

based colorimetric (cAMP, ERK phosphorylation) and ﬂuorometric (free cytosolic Ca²⁺) assays as per manufacturer’s instructions

(Abcam). All pharmacological TLR2 antagonism studies (Liang et al., 2011, 2013) described in this paper, were done at 10 mM ﬁnal

concentration of SpaB for 4 hours at 37^ꢀC. The number of live cells were estimated at the end of every experiment or assay using the

trypan blue reagent on a TC20 automated cell counter by as per manufacturer’s protocol (Bio-Rad), to ensure that the pharmaco-

logical treatments were not affecting cell viability.

Western blot analysis

Cell or tissue lysates were prepared as described earlier (Joshi et al., 2018), the protein concentrations from them were estimated

using the Pierce BCA Protein Assay kit (Thermo Fisher Scientiﬁc), and the western blot analysis was done using previously reported

protocols (Rajendran et al., 2020; Joshi et al., 2018). Brieﬂy, 50 mg protein lysate was resolved on a 12% SDS-PAGE gel and trans-

ferred onto a PVDF membrane (GE Healthcare) (60 mA, 12 h, 4^ꢀC). Post-transfer, the membrane was blocked with 5% (w/v) skimmed

milk in 1X phosphate buffered saline containing 0.1% (w/v) Tween-20 (PBST) (blocking buffer) for 1 h at room temperature (20 – 25^ꢀC),

washed with PBST (x 3 times) and subsequently probed with the primary antibody (dilution 1:1000) in blocking buffer for 12 – 14 h at

4^ꢀC. Following this, the membrane was washed with PBST (x 3 times), and incubated with a horse-radish peroxidase (HRP)

conjugated anti-rabbit IgG (H + L) (goat, Thermo Fisher Scientiﬁc, 31,460, 1:10,000) secondary antibody for 1 h at room temperature

(20 – 25^ꢀC). Finally, the membrane was washed with PBST (x 3 times), and the signal was visualized with SuperSignal West Pico

Chemiluminescent substrate (Thermo Fisher Scientiﬁc) on a Syngene G-Box Chemi-XRQ gel documentation system. For the phos-

phoprotein immunoblotting, the cells and tissues were lysed in buffer containing phosphatase inhibitors (1 mM sodium orthovana-

date, 1 mM sodium pyrophosphate, 5 mM sodium ﬂuoride) to maintain intact protein phosphorylation at serine, threonine and/or

tyrosine protein residues. A complete list of all the primary antibodies used in this study can be found in the Key resources table.

ELISA assays

All quantitative pro-inﬂammatory cytokine measurements (TNF-a, IL-6) were done using single analyte ELISA assays from R&D Sys-

tems as per manufacturer’s instructions. For measuring intracellular cAMP, cytosolic free Ca²⁺and phosphorylated ERK, ELISA kits

were purchased from Abcam, and the assays were performed as per manufacturer’s instructions. A complete list of all the ELISA kits

used in this study can be found in the Key resources table.

RT-PCR

The total RNA was isolated from mouse PPM using the RiboZol^TMRNA extraction reagent (Amresco), and the ﬁrst strand cDNA was

synthesized from the isolated total RNA using the ThermoScript^TMRT-PCR System (Invitrogen) as per manufacturer’s protocol. The

primers and the fragment size for the gene products analyzed in this study are listed in Table S1, and the gel for the RT-PCR exper-

iment is presented in Figure S3B.

Immunohistochemical analysis

The immunohistochemical analysis were done using a protocol recently published by us with minor modiﬁcations (Singh et al., 2020).

Brieﬂy, mice were deeply anaesthetized using isoﬂurane and perfused ﬁrst with cold phosphate buffer saline (PBS), and then with 4%

(w/v) paraformaldehyde (PFA) in PBS. Brains were dissected and post-ﬁxed in 4% (w/v) PFA in PBS overnight at 4^ꢀC and then trans-

ferred into 30% (w/v) sucrose in PBS until brains sank to bottom of the tubes (ꢂ3 days). Coronal 25 mm cerebellar sections were cut on

a freezing microtome (Leica CM1950), maintained at – 30^ꢀC and 1 out of every 6 sections was collected for assessing microglial acti-

vation. Endogenous peroxidases were inactivated with 3% (v/v) hydrogen peroxide in PBS for 15 min. Thereafter the sections were

washed twice with 1% bovine serum albumin (BSA) (HiMedia) in PBS, and permeabilized with 0.1% (w/v) Triton-X 100 and 0.5% (w/v)

BSA in PBS at room temperature (20 – 25^ꢀC) for 45 mins. The sections were then incubated with primary monoclonal anti-Iba1 anti-

body [rabbit, Abcam, Cat: ab178846, 1:1000] in 1% (w/v) BSA in PBS overnight at 4^ꢀC. Following overnight incubation, the sections

were washed three times with 0.5% (w/v) BSA in PBS and incubated with secondary antibody, biotinylated anti-rabbit [horse, Vector

labs, BP-1100, 1:500] in 0.5% (w/v) BSA in PBS for 1 h at room temperature (20 – 25^ꢀC). Following this treatment, the sections were

washed twice in 0.5% (w/v) BSA in PBS, incubated with ABC Elite Vectastatin (Vector labs, Cat: PK-6100) for 1 h at room temperature

(20 – 25^ꢀC) and subsequently washed twice with excess PBS. Finally, the sections were stained with Immpact DAB (Vector labs, Cat:

SK-4105) for 3 min in the dark, transferred to excess PBS and mounted in VectaMount (Vector labs, Cat: H-5000) after drying. Acti-

vated microglial cells (area >200 mm²) in matching cerebellar sections were quantiﬁed using the ImageJ software (NIH) (Rueden et al.,

2017; Schindelin et al., 2015) as per previously reported protocols (Blankman et al., 2013; Singh et al., 2020).

Histamine estimation from mast cells

The PCMCs were isolated from the peritoneal cavity of mice (10 -12 weeks of age) and cultured using a previously reported protocol

(Meurer et al., 2016) with typical yields of ꢂ (1 x 10⁶) PCMCs per mouse. Brieﬂy, cells from the mice peritoneal cavity were harvested

by lavage, using 10 mL cold sterile DBPS, and re-suspended in RPMI 1640 medium supplemented with 10% (v/v) FBS, antibiotics

Cell Chemical Biology 28, 1–11.e1–e6, June 17, 2021 e4

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

(1% (v/v) penicillin-streptomycin (MP Biomedicals)), 50 mM b-mercaptoethanol, 10 ng/mL recombinant mouse interleukin-3, and

30 ng/mL mouse stem cell factor at 37^ꢀC with 5% (v/v) CO₂. The cells were plated initially at a density of 0.5 x 10⁶cells/mL, and

48 hours post-plating, the non-adherent cells were removed by replacing the spent media with fresh media containing all the afore-

mentioned components. The adherent cells were allowed to grow and differentiate into PCMCs over the next 10 – 15 days with fresh

media supplementation (spent media was not discarded) on the 6^thday, and the purity of the PCMCs was assessed by Kimura stain-

ing (Kimura et al., 1973). Live PCMCs were estimated on a TC20 automated cell counter using trypan blue reagent as per manufac-

turer’s protocol (Bio-Rad) and used at the required cell density as per the desired assay condition. To assess the extent of lyso-PS

induced degranulation of PCMCs, histamine release from these cells was measured using an established LC-MS method (Chima-

lakonda et al., 2015), which we validated using an authentic histamine standard (Figure S7). Towards this, PCMCs were ﬁrst pelleted

(1400g for 5 min) and subsequently re-suspended in Hank’s Balanced Salt Solution (HBSS) containing 1% (w/v) heat inactivated FBS

at a concentration of 5 x 10⁶cells/mL. All assays were performed in a 96 well plates with ‘‘V’’ bottoms (Tarsons) to a ﬁnal volume of

100 mL at 37^ꢀC with 5% (v/v) CO₂. The vehicle (DMSO) or lipids (commercial lyso-PSs or (R)-Me-lyso-PSs or (S)-Me-lyso-PSs; ranging

from 1 nM – 100 mM concentration in the assay) were added to individual wells with or without Con-A (1 mg/mL ﬁnal assay concen-

tration) to make volume to 90 mL in HBSS. To this, 10 mL of the PCMC cell stock (5 x 10⁶cells/mL) pre-incubated at 37^ꢀC with 5% (v/v)

CO₂was added such that each assay well ﬁnally contained 5 x 10³cells, and the assay allowed to proceed for 30 mins at 37^ꢀC with

5% (v/v) CO₂. At this point, to quench the assay, the cells were pelleted by centrifugation (1200g for 5 mins), and 50 mL of the media

was collected for estimation of secreted histamine levels. To this 50 mL of media, 150 mL of LC-MS grade acetonitrile (Sigma-Aldrich)

was added, the mixture was vortexed and stored at – 80^ꢀC till the histamine estimations were performed by LC-MS. The pellet

cells were re-suspended in the remaining 50 mL media by vigorous vortexing and lysed by sonication in a water bath at 37^ꢀC.

The cell lysates were then added to 150 mL of MS grade acetonitrile (Sigma-Aldrich), the mixture was vigorously vortexed and

stored at – 80^ꢀC till the histamine estimations were performed. The histamine release from PCMCs was calculated in accordance

to a formula reported for this LC-MS method (Chimalakonda et al., 2015). For measuring circulating serum histamine concentrations,

1 volume of mouse serum (typically 200 – 300 mL) obtained from whole blood of wild type or ABHD12 knockout mice (12 – 14 weeks

of age), was mixed with 2 volumes (400 – 600 mL) MS grade acetonitrile (Sigma-Aldrich), the mixture was centrifuged at 2000g for

5 mins to pellet the protein debris, and the supernatant was collected, dried, and re-suspended in 50 mL MS grade acetonitrile

(Sigma-Aldrich), and assessed thereafter by LC-MS as per a previously reported protocol (Chimalakonda et al., 2015). The

quantitation for serum histamine levels, was done using a standard calibration curve for histamine (5 – 5000 ng/mL serum). All

histamine measurements were done using an established LC-MS method (Chimalakonda et al., 2015) in the positive ion

mode using high resolution multiple reaction monitoring (MRM-HR) analysis on a Sciex X500R quadrupole time-of ﬂight

˚

(QTOF) mass spectrometer ﬁtted with an Exion UHPLC system using a Kinetex 2.6 mm HILIC column with 100 A particle

size, 150 mm length and 3 mm internal diameter (Phenomenex). The MRM-HR mass spectrometry parameters for measuring

histamine are: precursor ion mass (Q1, M + H⁺) = 112.0869, product ion mass (Q3, M + H⁺) = 95.0604, declustering potential =

100 V, entrance potential = 10 V, collision energy = 50 V, and collision exit potential = 10 V. Complete details of the sensitivity

and dynamic range of the LC-MS method for histamine estimation, and the co-elution studies with a histamine standard can be found

in Figure S7.

Lyso-PS measurements

The organic extractions to enrich lyso-PS lipids, and all the LC-MS based quantitative lyso-PS measurements (secreted lyso-PS con-

centrations from PCMCs (Figure 6C), and the serum lyso-PS levels (Figure 6E)) were performed using an established and previously

reported protocols by us (Kelkar et al., 2019; Singh et al., 2020; Pathak et al., 2018). Brieﬂy, to 1 mL of serum free spent media from

PCMCs (Figure 6C) or mouse serum diluted to 1 mL using DPBS (Figure 6E), 3 mL of 2:1 (v/v) chloroform: methanol containing 1 nmol

of C17:1 lyso-PS (internal standard) was added. This mixture was vortexed, and then centrifuged at 2800g for 15 mins at 4^ꢀC to sepa-

rate the organic and aqueous layer. The organic layer (bottom) was collected, and the aqueous layer was acidiﬁed with 20 mL MS

grade formic acid. The acidiﬁed aqueous phase was vortexed, and the volume was subsequently adjusted to 4 mL using chloroform.

This mixture was vortexed, and centrifuged at 2800g for 15 mins at 4^ꢀC and organic layer (bottom) was collected. The organic layers

from both the extraction steps were pooled and dried under a stream of nitrogen gas at room temperature. The dried lipid extract was

re-suspended in 150 mL of 2:1 (v/v) chloroform: methanol and 20 mL of it was injected into the LC-MS for lyso-PS measurements. Area

under the curve for detected lyso-PS species was normalized to area under the curve for C17:1 lyso-PS (internal standard) and ab-

solute lyso-PS concentrations were obtained by normalized either to number of live cells (Figure 6C) or volume of serum (Figure 6E).

The measurements of lyso-PS (product) formed following porcine liver carboxylesterase (0.1 U, 15 mins, 37^ꢀC) or immune cell lysate

(ABHD12-null PPM or PCMC, 20 mg, 15 mins, 37^ꢀC) treatment of Me-lyso-PS (100 mg) was also done using the same reported pro-

tocols (Singh et al., 2020; Kelkar et al., 2019; Joshi et al., 2018).

Synthesis and compound characterization

Complete details of synthesis of all the compounds reported in this paper, their chemical/analytical characterization and all the

spectral data associated with these compounds can be found in the Data S1 associated with this paper.

e5 Cell Chemical Biology 28, 1–11.e1–e6, June 17, 2021

Please cite this article in press as: Khandelwal et al., Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs, Cell

Chemical Biology (2021), https://doi.org/10.1016/j.chembiol.2021.01.008

ll

Article

QUANTIFICATION AND STATISTICAL ANALYSIS

All bar graphs or data points represent mean of the experimental data, and the error bars represent the standard deviation from the

mean, as indicated in the ﬁgure legends. N denotes the exact number of biological replicates per experimental group, unless stated

otherwise in the ﬁgure legends. All statistical analyses were performed using the GraphPad Prism 7 (version 7.0e) software for Mac

OS. The Student’s two-tailed unpaired parametric t-test with a 95% conﬁdence interval using a two-stage linear step up procedure of

Benjamini, Krieger and Yekutieli (Benjamini et al., 2006) was used to determine the statistical signiﬁcance between the different study

groups in Figures 2B, 6C, and 6E, and a p value of <0.05 was considered statistical signiﬁcant for all the assays reported in this study.

p value annotations used in this study: *< 0.05, **< 0.01, and ***< 0.001.

Cell Chemical Biology 28, 1–11.e1–e6, June 17, 2021 e6

Article Doi

Fatty acid chain length drives lysophosphatidylserine-dependent immunological outputs

DOI: 10.1016/j.chembiol.2021.01.008

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.chembiol.2021.01.008

Full text of DOI:10.1016/j.chembiol.2021.01.008

Products guided by the article

R&D Labs maybe for 51826-99-4

Relevant to this article

Hot Product