Human phospholipases A2: a functional and evolutionary analysis

Phospholipases A2 (PLA2) are capable of hydrolyzing the sn-2 position of glycerophospholipids to release fatty acids and lysophospholipids. The PLA2 superfamily enzymes are widespread and present in most mammalian cells and tissues, regulating metabolism, remodeling the membrane and maintaining its homeostasis, producing lipid mediators and activating inflammatory reactions, so disruption of PLA2-regulated lipid metabolism often leads to various diseases. In this study, 29 PLA2 genes in the human genome were systematically collected and described based on literature and sequence analyses. Localization of the PLA2 genes in human genome showed they are placed on 12 human chromosomes, some of them forming clusters. Their RVI scores estimating gene tolerance to the mutations that accumulate in the human population demonstrated that the G4-type PLA2 genes belonging to one of the two largest clusters (4 genes) were most tolerant. On the contrary, the genes encoding G6-type PLA2s (G6B, G6F, G6C, G6A) localized outside the clusters had a reduced tolerance to mutations. Analysis of the association between PLA2 genes and human diseases found in the literature showed 24 such genes were associated with 119 diseases belonging to 18 groups, so in total 229 disease/PLA2 gene relationships were described to reveal that G4, G2 and G7-type PLA2 proteins were involved in the largest number of diseases if compared to other PLA2 types. Three groups of diseases turned out to be associated with the greatest number of PLA2 types: neoplasms, circulatory and endocrine system diseases. Phylogenetic analysis showed that a common origin can be established only for secretory PLA2s (G1, G2, G3, G5, G10 and G12). The remaining PLA2 types (G4, G6, G7, G8, G15 and G16) could be considered evolutionarily independent. Our study has found that the genes most tolerant to PLA2 mutations in humans (G4, G2, and G7 types) belong to the largest number of disease groups.

The PLA2 family is one being most extensively studied, which reflects their biological importance. They hydrolyze the ester bond of membrane phospholipids from the sn2 position, and, under natural conditions, their sn2 positions often contain polyunsaturated fatty acids, which, when re leased, can be metabolized to form various eicosanoids and their associated biologically active lipid mediators (Aloulou et al., 2018).
Assigning a PLA2 to a certain group (type) is based on the experimental determination of their catalytic mechanisms, cellular localization, evolutionary and structural features. Note that most of these lipolytic enzymes share no struc tural similarity and have different regulatory and catalytic mechanisms (Aloulou et al., 2018).
Each of the sixteen PLA2 types is involved in lipid me tabolism and disease development mechanisms of different kind, so PLA2s are believed to be promising therapeutic targets for a number of diseases (Aloulou et al., 2018). In this respect, there is a huge interest in the pharmaceutical industry for development of selective and effective inhibitors for each of these PLA2 types (Aloulou et al., 2018).
Describing protein functions is known to include, on the one hand, the molecular function, and, on the other, the function at the level of the vital activity of a cell or a whole organism (Karp, 2000). PLA2s have been fairly well studied in terms of their molecular functioning, however, their role in the vital processes of a cell and a whole organism remains poorly understood.
The objective of the present study was to analyze the cha racteristics of various human PLA2 types in the context of the available data on their association with various diseases. To do so, the PLA2s' proteinsequence domain organization, gene distribution in the genome, mutability characteristics as well as their phylogenetic relationships with the PLs of other organisms were analyzed.

Materials and methods
Sampling of human and animal PLs. The human PLA2 protein sequences were taken from Huang Q. et al. (2015), and since not all known human PLA2s were described in this paper (Dennis et al., 2011), the missing sequences were identified in the NCBI database by their names and identi fiers as per Dennis et al. (2011), using the GRCh38.p14 human genome assembly.
The primary structures of human PLA2s were characte rized by the presence of domains, active sites, and signal peptides using the published data. To search for the PLA2s' homologues in animals, the BLASTP program (E-value ≤ 1) was employed with the human PLA2protein sequences used as a query. The homologues were searched for among the protein sequences of the organisms representing various taxa, for their list see Suppl. Material 1.
Functional analysis of the PLA2s. To estimate the degree of the PLA2 genes' evolutionary conservation, the Residual Variance Index Score (RVIS) (Petrovski et al., 2013) was applied. The scoring enables one to assess a gene's tolerance to the mutations that accumulate in the human population, so the score is calculated based on the allele frequency information presented in the entire human exome sequence (data set NHLBIESP6500 from EVS v.0.0.14: https://evs. gs.washington.edu/EVS/). The score allows ranking genes by the number of observed nucleotide variations, taking into account the relative proportion of neutral substitutions that are observed for a gene under study. If negative, its value indicates low gene variability (i. e., its sequence is less tolerant to the accumulated mutations found in genes with a more important function), and if positive, it shows a higher gene variability (i. e., its sequence is more tolerant to nucleotide substitutions).
The DAVID service (Huang D.W. et al., 2009) was em ployed to identify the biological processes involving PLA2s. The service allows one to identify the terms from the Gene Ontology, INTERPRO and KEGG Pathway databases, over represented in the annotations of the genes from an analyzed sample in comparison with the annotations of all genes in a human body. In our case, such a sample was a sample of human PLA2 genes.
Searching for PLA2/disease associations. The search for the articles describing the relationship between human diseases and PLA2protein activity was carried out in the PubMed and Google Scholar databases using such queries as "disease/patients/pathology/name of a specific disease (e. g., lung cancer or schizophrenia) + PLA2/phospholipase A2/ name of a specific PLA2 (e. g., pla2g1b, pla2g2a)". Informa tion was also taken from the reviews on PLA2 involvement in various diseases.
The found articles tracked information about the asso ciation of a person's disease and the activity/expression of a specific PLA2. For example, such information included reports about the patients who had significantly reduced/ increased expression or activity of a certain PLA2 compared to healthy people; data that PLA2 gene mutation enhanced/ weakened the severity of a disease; data that the mechanism enabling a PLA2 to influence the course of the disease had been established. To classify diseases in this study, the In ternational Classification of Diseases (ICD-10 available at https://icd.who.int/browse10/2019/en; in Russian at https:// mkb10.com) (Hirsch et al., 2016) was used.
Based on the information about the relationship between a disease and PLA2 involvement in it, a data table was formed, whose rows listed human disease types, and the columns -PLA2 types. If the table's cell had a value of 1, it meant this PLA2 type was involved/associated with the disease. To build this table, a Python script had been written, linking the name of a disease to its ICD10 code.
As the next step, a hierarchical clustering of the human PLA2 types was performed according to the degree they were associated with various diseases. To do so, for different types of phospholipases, the degrees of their participation in the diseases from the abovementioned table were compared for different PLA2 types, using the Euclidean distance as a measure of similarity and the unweighted pair group method with arithmetic mean (UPGMA) -for clustering. In the same way, the diseases were clustered based on the degree of their association with different PLA2 types.
Multiple sequence alignment and protein phylogeny reconstruction. Multiple alignment of homologous PLA2 sequences was performed using the PROMALS (Pei, Grishin, 2007) and MAFFT (Katoh, Toh, 2010) software. The search for proteins for alignment and, accordingly, the alignment of protein sequences were carried out only in the PL domain. The phylogenetic tree was reconstructed using the maximum likelihood method and the IQ-TREE software (v.8.2.4, see (Nguyen, 2015)) with an optimal WAG + R6 model chosen.

Structural and functional characteristics of the human PLA2s
The features of the structural organization of the various types of the human PLA2s are shown in Figure 2. The proteins' properties (substrates, activity, mass, catalytic residues, etc.) is given in Suppl. Material 3.
Phospholipids served as substrates for sPLA2 enzymes. In all cases, these were either phosphatidylcholine (PC) or phosphatidylethanolamine (PE), except for pla2g12a, whose substrate was phosphatidylglycerol (PG) but not PC or PE. Some sPLA2s also had PG and phosphatidylserine (PS) as substrates. The human pla2g12b protein was catalytically inactive (see Suppl. Material 3 and caption to Fig. 2).
Cytosolic phospholipases A2 (cPLA2). The cPLA2s were represented by the G4 type of PLA2 that included six human proteins: pla2g4(a-f ). The mass of cPLA2 proteins varied from 541 aa (pla2g4с) to 1012 aa (pla2g4b) (see AdPLA2  The red rectangles mark the active sites, the blue one in the pla2g12b sequence denotes H (histidine) replaced by L (leucine) at position 115 of the protein that kills its catalytic activity (Guan et al., 2011). CBL is a Сa 2+ binding loop. cNMP is a domain biding cyclic nucleotides (cAMP or cGMP). The pancreatic loop is sPLA2G1B of unique five-amino-acid extension. The Cap is a domain found in PLA2G4A that opens/closes an active site for PL substrate modeling. The drawing was adopted from (Kudo, Murakami, 2002;Dennis et al., 2011). Fig. 2). In the proteins, the catalytic domains were located in the Cterminus of the sequences and contained a conserva tive Ser/Asp catalytic dyad (see Suppl. Material 3; Fig. 2). As sPLA2s, cPLA2s are calciumdependent PLA2, so they also had a calcium binding domain closer to the Nterminus (Dennis et al., 2011).

ЭВОЛЮЦИОННАЯ КОМПЬЮТЕРНАЯ БИОЛОГИЯ / EVOLUTIONARY COMPUTATIONAL BIOLOGY
In the G4 type proteins (cPLA2/PLA2G4), as well as in sPLA2 proteins, PLA2 activity was observed if their sub strates were either PC or PE. The pla2g4a protein addition ally had phosphatidylinositol (PI) as a substrate, while the pla2g4c protein had PC, but its specificity for PE was not demonstrated (see Suppl. Material 3).
Calcium-independent phospholipases A2 (iPLA2). The iPLA2s included only G6type PLA2s, their length varying from 253 to 1365 aa. In iPLA2 proteins, the catalytic do mains were located closer to the Cterminus in pla2g(a, c) and closer to the Nterminus in pla2g(d-f ) (see Fig. 2). As that of cPLA2s, the protein's catalytic domain contained a conservative catalytic Ser/Asp dyad (see Suppl. Material 3; Fig. 2). As reflected in their name, iPLA2 catalytic activity was independent of Ca 2+ presence, and, unlike the sPLA2s and cPLA2s, they did not have a Ca 2+ binding domain. The pla2g6a protein had a region containing 7 ankyrin repeats closer to the Nterminus. This motif is involved in protein protein interactions, allowing intensive binding to membrane proteins (Filkin et al., 2020). In the pla2g6c protein, closer to the Cterminus were three cNMP sites (sitebinding cyclic nucleotides) (see Fig. 2).
The g6(a-f ) proteins used PC as substrates for PLA2 reactions; in the case of the pla2g6b protein, it could have PE in addition to PC. In addition to PLA2, these enzymes could also exhibit other activities such as TGhydrolase, lysophospholipase, PLA1 (for pla2g6b) and other ones (see Suppl. Material 3).
Platelet activating factor acetyl hydrolase (PAF-AH or Lp-PLA2). The PAFAHs included G7 and G8 PLA2s that modulated the activity of a platelet activating factor (PAF), a potent phospholipid inflammation mediator involved in inflammation, platelet aggregation and anaphylactic shock pathogenesis (Shimizu, 2009). The length of PAFAH proteins was 441 and 392 aa for g7a and g7b, and 229 and 231 aa -for g8a and g8b, respectively (see Fig. 2). In the proteins, the catalytic domain occupied almost the entire sequence and contained a conserved Ser/His/Asp catalytic triad (see Suppl. Material 3; Fig. 2). They were independent of Ca 2+ and had no Ca 2+ binding domain (see Fig. 2).
Proteins pla2g7(a, b) and pla2g8(a, b) were able to hy drolyze a phospholipid platelet activating factor (PAF) into a lysoPAF. At the same time, the pla2g7a protein possessed both PLA2 and PLA1 activities and could use as a substrate both PC and oxPC. The pla2g7b protein showed a PLA2 activity (see Suppl. Material 3).
Lysosomal phospholipases A2 (LPLA2). The LPLA2s (G15type PLA2s) were represented by a single pla2g15 protein of 412 aa in length (see Fig. 2). The protein's catalytic domain was located in the central region of the sequence and contained a conservative Ser/His/Asp catalytic triad (see Suppl. Material 3; Fig. 2). The lpla2 protein was independent of Ca 2+ and had no Ca 2+ binding domain.
The pla2g15/lpla2 protein of type G15 had PLA2 and PLA1 activities, whose substrates being PC, PE and PS. Also, pla2g15 was capable of acylceramide synthase activity through C1 ceramide (see Suppl. Material 3).
Adipocyte phospholipases A2 (AdPLA2). The AdPLA2s (G16type PLA2s) were represented by a single pla2g16 pro tein of 193 aa in length (see Fig. 2). In pla2g16, the catalytic domain was located in the central region of the sequence and, like in LPLA2s contained a conservative Ser/His/Asp catalytic triad (see Suppl. Material 3; Fig. 2). The adpla2 protein was independent of Ca 2+ and, as iPLA2, PAFAH, LPLA2, had no Ca 2+ binding domain (see Fig. 2).
The protein had PLA2 and PLA1 activities through PC and PE substrates and a NacylPE acyltransferase activity, through diacyl PE (see Suppl. Material 3).

PLA2-gene localization in human genome
The localization of the PLA2 genes in the human genome (version GRCh38.p14) is shown in Figure 3. The genes were absent on the 2, 3, 5, 8, 9, 13, 14, 17, 18, 20, 21st and Y chromosomes. The 4, 6, 7, 10, 12th and X chromosomes contained one PLA2 gene; chromosomes 11, 16 -two PLA2 genes; chromosomes 19, 22 -three PLA2 genes. In the 15th chromosome of the four genes (G4B, G4E, G4D, G4F ) formed a 0.3 Mb cluster at the 43 Mb position. On chromosome 1, in addition to a single G4A gene (at 188 Mb), at the 20 Mb position was a cluster of six genes (G2E, G2A, G5, G2D, G2F, G2C ) of 0.11 Mb in size. It is noteworthy that, excluding the genes of these two clusters, all other genes were isolated from one another at a distance of at least 6 Mb. Moreover, while the G4type PLA2 genes (G4B, G4E, G4D, G4F ) were located in the abovementioned cluster on chromosome 15, the other two genes of this type (G4A and G4C) were isolated on chromosomes 1 and 19, respectively, so all PLA2 genes of type G2 (G2A, G2C, G2D, G2E, G2F ) were located in a cluster on chromosome 1, but, together with them, this cluster included the G5 gene. The G6type PLA2 genes were located: G6B -on chromosome 7, G6E -on chromosome 11, G6C -on chromosome 19, G6A and G6Don chromosome 22, and G6F -on the X chromosome. Figure 4 displays RVI-score distribution for human PLA2s. On the left of the graph are PLA2s whose score is above zero, so these are genes that contain a relatively large num ber of mutations and are tolerant to them. To the right are PLA2s whose score is below zero, so they are less tolerant to mutations. The genes of G16, G1, G12 (PLA2G12B), G4 (PLA2G4A), G5, G15, and G6 types had a negative RVI score (see Fig. 4). Of these PLA2s, three were secreted G1, G12 and G5 types as well as calciumindependent (type G6), cytosolic (G4), lysosomal (G15), and adipocyte (G16) PLA2 genes. Interestingly, four of the six PLA2 genes of G6 type Human phospholipases A2: a functional and evolutionary analysis G1B(g ) 1b
The Y-axis's right part on corresponds to the proportion of genes in the human genome (in %) whose RVI score is less than that for a particular gene (bar). These percentile values are marked on the graph as orange lines. The columns of other colors mark PLA2s of different types: cPLA2 (dark blue), sPLA2 (green), iPLA2 (pink), PAF-AH (red), adPLA2 (pla2g16; black), LPLA2 (pla2g15; blue).
The most tolerant to mutations were the cPLA2 genes of G4 type. Five out of the six genes of this kind had a positive RVI score and only one (PLA2G4A) -a score less than 0 (RVIS = -0.25). The most mutation-tolerant genes in this group (PLA2G4(B, D-F )

Fig. 5.
Signaling pathways and biological processes from the KEGG Pathway database that were detected by the DAVID service as significantly associated with the found PLA2 genes.
Along the Y-axis are terms describing the signaling pathways and biological processes. In brackets, after each term, the false discovery rate value (FDR or expected proportion of false rejections) is given. The X-axis plots the number of PLA2 genes associated with each term.
on chromosome 1. For five human PLA2s (PLA2G2C, PLA2G7B, PLA2G8A, PLA2G8B, PLA2G10), the EVS server did not contain any genevariability data to calculate their RVI score (see Materials and methods section), so they were excluded from the graph (see Fig. 4).
Human PLA2 relationship to the biological processes and signaling pathways from the KEGG Pathway database Figure 5 demonstrates the results of a functional analysis of the found PLA2 genes performed using the DAVID service. It turned out that the most significant (based on the number of PLA2 genes associated with it) was the Ras signaling pathway that was involved in carcinogenesis. Another most used term was the VEGF pathway associated with a vascular endothelial growth factor. The diseases associated with this pathway also tended to be associated with the development of such tumors as breast cancer, glioma, melanoma, etc. (Takahashi, Shibuya, 2005). Thus, the data have shown that PLA2s are significantly associated with carcinogenesis.  Figure 6. Eighteen disease groups, their names and ICD10 codes (in parentheses) are the rows of the clustering diagram. The bars in the diagram correspond to the 12 types of human PLA2s. Most PLA2 groups were associated with neoplasms (ICD10 code: C00-D48; 9 groups in 12); diseases of the circulatory system (I00-I99; 8 in 12); diseases of the endo crine system (E00-E90; 7 in 12); diseases of the eye and adnexa (H00-H59; 6 in 12). The smallest number of PLA2 groups was associated with congenital anomalies (Q00-Q99; only one G7type PLA2); symptoms, signs and abnormali ties (R00-R99; one G6-type PLA2); certain infectious and parasitic diseases (A00-B99; only G2 and G7type PLA2s).

PLA2-associated diseases
It is interesting to note that out of considered PLA2 types, the following had most associations with diseases: G7 was associated with 15 disease groups out of the 18 presented in Figure 6; G2 -with 13 groups; G4 -with 12 groups. The least represented in the disease groups were: G8 associated only with diseases of the genitourinary system (N00-N99); G15 -only with diseases of the circulatory system (I00-I99); G12 -only with mental and behavioral disorders (F00-F99) and with diseases of the eye and adnexa (H00-H59); G16only with neoplasms (C00-D48) and with diseases of the endocrine system (E00-E90).
The horizontal clustering of the PLA2s demonstrated their division into three clusters (see Fig. 6). The first contained the G4, G2, G7 types and the genes were involved in a large number of the human diseases analyzed. The second cluster Human phospholipases A2: a functional and evolutionary analysis The cells are colored black when links between the genes of a PLA2 group and the diseases from the presented disease groups have been revealed. The white color marks the cases when no gene-disease links have been identified. Diseases of the genitourinary system (N00-N99) Certain conditions originating in the perinatal period (P00-P96) Diseases of the eye and adnexa (H00-H59) Neoplasms (C00-D48) Diseases of the endocrine system (E00-E90) Diseases of the circulatory system (I00-I99) Diseases of the respiratory system (J00-J99) Diseases of the musculoskeletal system and connective tissue (M00-M99) Mental and behavioral disorders (F00-F99) Symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified (R00-R99) Diseases of the blood and blood-forming organs and certain disorders involving the immune mechanism (D00-D59) Certain infectious and parasitic diseases (A00-B99)
It is noteworthy that of the twelve studied PLA2 genes of G4, G2, G7 types, eleven had a high level of mutation tolerance (RVIS) and only one, PLA2G4A, had a mode rately low level of mutation tolerance (RVIS = -0.25) (see Fig. 4), given that these types are involved in the greatest number of diseases (see Fig. 6). At the same time, of the seven studied PLA2 genes of types G6 and G15, five had the lowest level of tolerance to mutations (RVIS) and only one, PLA6G6D, had a relatively high level of tolerance to muta tions (RVIS = 0.85) (see Fig. 4), given that these types belong to the cluster associated with the least number of diseases (see Fig. 6). This suggests a possible positive relationship between the number of diseases in which a PLA2 is involved and the gene's mutation tolerance (RVI score). However, calculating the correlation coefficient by the χ 2 method did not reveal a significant correlation between these values, so, in this case, we can only speak of an unreliable trend.
At the same time, Petrovski et al. (2013) studying a sample of the genes associated with Mendelian (monogenic) diseases, demonstrated that they had a low tolerance (RVI score) compared to other human genes. The authors sug gested that a negative RVI score indicated the presence of purifying selection, and a positive one -either the absence of purifying selection, or even the presence of some form of balanced or positive selection.
The tendency towards an increased RVI score in the PLA2 genes involved in a greater number of diseases, may be due to the fact that, when considering expression data, a signal from a set of identified differentially-expressed genes can be significantly contaminated by the noisy produced by random genes. The appearance of such random genes can be associated as with the features of an applied technique (Hatfield et al., 2003) as with the fact that any perturbation in the cell and organism (e. g., a disease) can induce nonspe cific effects on gene expression (e. g., stress response genes activation, apoptosis, necrosis, etc.) (Leuner et al., 2007;Turkmen, 2017). Therefore, when estimating the number of associations between a PLA2 and a disease, both large and small numbers of associations must be interpreted with some caution.

PLA2 evolution
Searching for homologous sequences in the protein databas es was employed to identify PLA2 sequences for 32 species (see Suppl. Material 1), including 13 vertebrates and 19 in vertebrates (see the Materials and methods section). Their identifiers and sequences are given in Suppl. Material 5.
To illustrate the similarity of PLA2 functional regions, a homology analysis of the catalytic domains of the hu man PLA2s was performed, and such a similarity between PLA2 domains of different types was only found among secretory PLA2s (Suppl. Material 6), in particular between the catalytic domains of G1, G2, G5, G10 types the Evalue varied from 2e-03 to 2e-38; and between G12 type proteins (g12a and g12b) it was equal to 4e-48. At the same time, no similarity (E-value ≥ 1) was detected between G3-type PLA2s (plag3) and all other sPLA2 proteins.
Respectively, there was no similarity between the human PLA2 domains of these three G6 subtypes. G7 (two proteins g7a and g7b) and G8 (two proteins g8a and g8b) types had similarities within these type of sequences: 7e-103 (G7) (Suppl. Material 9) and 3e-102 (G8) (Suppl. Material 10). For the remaining two types, no comparison was made, since in human, they included only one protein each.
To reconstruct the phylogeny of the PLA2 proteins, a homology and multiple sequence alignment analysis for the proteins of different PLA2 types had been initially per formed. It showed that the proteins of the secreted sPLA2 The type names (clusters on the tree) are given in blue text that describes which taxa are represented in each cluster. The red texts and squares highlight human PLA2 proteins. Their two types of bootstrap support are shown next to the tree nodes separated by a slash: ultrafast bootstrap (UFBoot)/bootstrap SH-aLRT. A textual description of the tree is given in Suppl. Material 11. Human phospholipases A2: a functional and evolutionary analysis group (G1-3, G5, G10, G12 types) had high or moderate homology (E-value ≤ 1) and qualitative alignment within the group. In contrast, the proteins of other PLA2 types (G4, G6, G7, G8, G15, G16) had very low homology (Evalue > 1) and poorly aligned as between themselves as with respect to the sPLA2 proteins. In this respect, sPLA2 phylogeny was reconstructed using the maximum likelihood method (Fig. 7).
The results of the phylogenetic analysis enabled us to assume that two successive divergences occurred in the common ancestors of multicellular invertebrates: first, the ancestral sPLA2 gene diverged into the G3/G12 and G1/ G2/G5/G10 ones, and then the G3/G12 gene diverged into the ancestral G3 and G12 genes.

Conclusion
The paper presents the results of analysis of the PLA2 fa mily in human and describes the structure and functions of 29 PLA2s belonging to 12 types: G1-8, G10, G12, G15, G16. Analysis of PLA2gene localizations in the human genome has demonstrated they present on 12 chromosomes and some of them form clusters, the two largest of them in clude, first, all G2-type PLA2 genes (G2A, G2C-F ) and the G5 gene, and second -G4 type PLA2 genes (G4B, G4D-F ).
The association between the PLA2s and human diseases as they described in the literature have also been analyzed. In total, 229 disease-PLA2 gene links have been found, so associations between 24 PLA2 genes and 119 diseases have been demonstrated. The PLA2 proteins of types G4, G2 and G7 have turned out to be involved in the greatest number of diseases if compared to the other types, whereas three groups of diseases have turned out to be associated with the largest number of PLA2 types: neoplasms, circulatory and endocrinesystem diseases.
RVI scoring of the genes' tolerance/intolerance muta tions has showed that the majority of genes of the G4 (G4B, G4C, G4D, G4E, G4F ) and G2 (G4D, G4E, G4F ) types, as well as the genes of the types represented by one G3 and G7 gene, were tolerant to mutations, whereas most genes of the G6 type (G6A-C, G6F ) as well as the types represented by a single gene (G5 and G15), turned out to be not tolerant. Here it should be noted that all the PLA2 types with predominance of genes tolerant to mutations, except for G3, have also been associated with the greatest number of diseases: G4 (12 disease groups), G2 (13), G7 (15), while all the PLA2 types intolerant to mutations have been asso ciated with a smaller number of disease groups: G6 (9 di sease groups), G6 (7), G6 (1), which suggests that higher tolerance to mutations in a particular human PLA2 gene is associated with its involvement in more diseases or disease groups.