Relationship between SPRY and B30.2 protein domains. Evolution of a component of immune defence?

Abstract

SPRY and B30.2 are homologous domains which can be identified in 11 protein families encoded in the human genome. These include cell surface receptors of the immunoglobulin super-family (BTNs), negative regulators of the JAK/STAT pathway (SOCS-box SSB1–4) and proteins encoded by the numerous TRIM genes. Collectively, proteins containing SPRY and B30.2 domains cover a wide range of functions, including regulation of cytokine signalling (SOCS), RNA metabolism (DDX1, hnRNPs), intracellular calcium release (RyR receptors), immunity to retroviruses (TRIM5alpha) as well as regulatory and developmental processes (HERC1, Ash2L). In order to clarify the evolutionary relationship between the two domains, we compiled a curated database of SPRY and B30.2-domain sequences. We show that while SPRY domains are evolutionarily ancient, B30.2 domains, found in BTN and TRIM proteins, are a more recent evolutionary adaptation, comprising the combination of SPRY with an additional domain, PRY. The combination of SPRY and PRY to produce B30.2 domains may have been selected and maintained as a component of immune defence.

Introduction

B30.2 was identified as a protein domain encoded by a single exon within some genes in the major histocompatibility complex (MHC) region of human chromosome 6p21.3.¹ The domain, of ∼170 amino acids, comprised the carboxy terminus of proteins belonging to three different protein families:² BTNs, which are receptor glycoproteins of the immunoglobulin superfamily; representatives of the large group of TRIM proteins, composed of the RING/B-box/coiled-coil (RBCC or tripartite motif) core with amino- and carboxy-terminal additions;³ and stonutoxin (STNX), a secreted venomous protein product of the stonefish Synanceia horrida.^4,5

Contemporaneous with the description of the B30.2 domain was that of the SPRY domain.⁶ Initially identified within ryanodine receptor (RyR) sequences from various mammalian species, sequence homology was detected with non-receptor tyrosine kinase spore lysis A (SplA) of Dictyostelium discoideum. The SPRY domain was subsequently identified in a total of 10 distinct protein families, including the human sequences DDX1, hnRNPs, HERC1 and RanBPM.⁶

There is some confusion over the nomenclature of the B30.2 and SPRY domains and the relationship between them. In sequence databases and in the literature, these terms are used interchangeably. In addition, the protein domain termed PRY appears to be associated with the SPRY domain in some cases. A comprehensive analysis of SPRY and PRY domain sequences has been produced (http://smart.embl-heidelberg.de/smart/),⁷ which includes the molecules with carboxy terminal B30.2 domains.

Starting with well-defined B30.2 domain protein sequences, we used a sensitive homology search based on hidden Markov Model profiles (HMMs) to identify all protein sequences containing the ∼170 amino acid B30.2 domain. Over 1800 sequences with B30.2, SPRY and PRY domains were obtained. We discuss the relationship between these protein domains in terms of their evolution, function and primary structure. Our database of sequences, arranged according to protein family and by organism, is available at http://www.genomesapiens.org/b302/browser.html.

Alignments, motifs and domain definitions

Sequences conforming to 16 B30.2 domains were used initially to provide a seed alignment to search NRDB (a database of protein sequences from which near-neighbour redundant sequences have been removed) using a single HMM based upon the alignment. This initial set was composed of amino acid sequences of well-defined B30.2 protein domains, consisting of BTNs, representatives of the TRIM family and stonutoxin (STNX), the only non-human sequence used (5 and Fig. 2a). Of the 166 sequences returned, 122 were confirmed by BLASTp searches to possess the B30.2 domain, being primarily more TRIM-like sequences. The B30.2 domain regions from these 122 sequences were each used to create an individual HMM using the SAM T-99 procedure.

Figure 2.

(a)Alignments of representative B30.2 domains used to seed the HMM analysis. These represent human BTN and TRIM B30.2 amino acid sequences and STNX, the only non-human sequence used. Three blocks of conserved residues are evident (underlined) giving the motifs used previously to characterize B30.2 domains.⁵ (b) Alignments of representative SPRY domains from human proteins. Two minimal motifs can be identified (underlined) which accounts for the amino acid similarity to B30.2 sequences detected by HMM analysis. (c) A comparison of conserved motifs in B30.2 and SPRY domains. The B30.2 domain is longer than SPRY and is composed of PRY and SPRY domain motifs combined. Amino acid identity between B30.2 and SPRY domains amounts to 9 conserved residues, comprising two minimal motifs, separated by up to 80 amino acids. B30.2 and SPRY are very distant homologues. Boxed residues in motif 2 highlight the block of three conserved aromatic residues.

This HMM set was used to search NRDB using an E-value cut-off of 0·01. A large number of B30.2, SPRY and PRY protein sequences were identified (1891) which, by 99% probability were of the same structural super-family as the initial B30.2 set. As can be seen from alignments of representative members of B30.2 and SPRY sequences (Fig. 2) B30.2 domain sequences are longer than SPRY and contain three blocks of conserved residues. Three sequence motifs, LDPD, WEVE and LDYE, have been used previously to define B30.2 sequences (5 and Fig. 2a). For the shorter SPRY sequences, shared protein motifs are not so obvious, although conserved amino acid residues can be detected (6 and Fig. 2b). Based on the consensus alignments shown in Fig. 2, we attempted to trace the relationship between SPRY and B30.2 domains (Fig. 2c). SPRY sequences are truncated at the amino-terminus in relation to B30.2 and they are missing motif 1 (LDPD). These additional amino acids are referred to as the PRY domain in sequence databases. This shows that the B30.2 domain is composed of a combination of PRY and SPRY domains. Motif 2 comprises five conserved residues with 3 aromatic amino acids. Motif 3 consists of four conserved residues. Thus amino acid conservation between B30.2 and SPRY domains amounts to nine conserved residues comprising two minimal motifs separated by up to 80 amino acids. Our HMM analysis detected SPRY sequences with E < 0·01, using B30.2 sequences as a start point. This confirms a distant evolutionary relationship in primary amino acid sequence between SPRY and B30.2 sequence segments. Within individual protein families with SPRY or B30.2 domains and across species as diverse as man and Drosophila, amino acid conservation is very high, such as in ryanodine receptors (RyR).

SPRY and B30.2 domains in 11 protein families

In order to identify redundant sequences and close homologues from the database, sequences were clustered with a percentage cut-off of 40% amino acid identity. Clustered sequences were grouped into different protein families according to the presence of additional protein domains, defined using the InterproScan procedure⁸ and by BLASTp.⁹ Eleven distinct protein families were identified in sequences from Homo sapiens, where the SPRY or B30.2 domain is found in combination with a diverse set of additional protein domains (Fig. 1). Sequences which could not be categorized using these criteria were placed in a separate ‘other’ group.

Figure 1.

Structure and function of SPRY and B30.2 protein families. Schematic representation of the domain architecture of 11 protein families with SPRY or B30.2 domains in the human genome. Representative amino acid sequences of human proteins were used and domain assignments predicted using BLASTp.⁹

A diverse set of molecules was identified with SPRY domains. The SPRY domain was found in various positions in relation to other protein domains within representative molecules and with more than one copy in RyR and FLJ14800. In the human genome, B30.2 domains are found only in BTN and TRIM molecules and always at the carboxy terminus. The B30.2 domain is encoded by a single exon and alternatively spliced transcripts lacking the domain have been described for both BTN and TRIM genes.

Features of SPRY and B30.2 containing proteins

Figure 1 lists the principal features of human proteins containing SPRY and B30.2 domains. The protein families cover a wide range of functions. Proteins with SPRY domains are involved in RNA metabolism (DDX1, hnRNPs), intracellular calcium release (RyR receptors), regulatory and developmental processes (HERC1, Ash2) and possibly in the regulation of cytokine signalling (SOCS box SS-B). TRIM molecules have been associated with some inherited diseases in man (pyrin/MID1), with ubiquitin protein ligase activity (MID1) and with innate immune resistance to retroviruses (TRIM5alpha). BTN molecules appear to be more specifically involved in regulation of the immune system (Fig. 1).

Evolutionary conservation of SPRY and B30.2 protein families in model organisms

We looked for conservation in model organisms of the SPRY and B30.2 domain protein families identified in human sequences. Table 1 shows the number of genes and conservation of the protein families in mouse, Drosophila, Caenorhabditis elegans and yeast. Of the 11 clearly identified protein families described in human sequences, all were identified in mouse, including TRIM proteins and representatives of the BTN set. Seven protein families were identified in Drosophila and C. elegans, while two homologues, RanBPM and Ash2, were positively identified in yeast.

Table 1.

Number of genes and conservation of SPRY and B30.2 protein families in model organisms. The presence of a homologous molecule is indicated by (✓). The absence of this molecule is indicated by (×).

		Conservation
Protein family	Homo sapiens# of genes	Mus	D. melanogaster	C. elegans	Yeast	Other
TRIM	60	✓ 43	×	×	×	Xenopus 4
BTN	9	✓ 4	×	×	×
SS-B	4	✓ 3	✓	✓	×	Anopheles
hnRNP	3	✓ 2	×	✓	×
RyR	3	✓ 2	✓	✓	×	Anopheles
RanBPM	2	✓ 2	✓	✓	✓	Xenopus
DDX1	1	✓ 1	✓	✓	×
HERC1	1	✓ 1	×	×	×
ASH2L	1	✓ 1	✓	✓	✓
FLJ14800	1	✓ 1	×	×	×
KIAA1972	1	✓ 1	✓	✓	×

This analysis indicates that SPRY domains are evolutionarily more ancient than B30.2, because examples of SPRY are found in animals, plants and fungi. B30.2 domain sequences on the other hand were identified only in human, mouse, chicken¹⁰ and Xenopus. Genes with B30.2 domains are more abundant due to an expansion in gene number of TRIM and, to a lesser extent, BTN genes in mammalian genomes. The genomes of human and mouse contain different numbers of TRIM and BTN genes so clear orthologues of some human genes are not all identifiable in the mouse genome.

B30.2 domains are only present so far in vertebrate species with an adaptive immune system and the number of genes with B30.2 domains varies in human and mouse genomes, species separated by ∼80 Mya. This plasticity in gene organization and recent evolution are features which are similar to some immune receptors. The gain/loss of these genes may be ongoing as a result of selective pressure.^11,12

TRIM5alpha interacts with human immunodeficiency virus-1 (HIV-1) via the B30.2 domain

TRIM5alpha has been shown to be a cellular restriction factor for infection by HIV-1.¹³ Taking advantage of differences in the cellular tropism shown by human and rhesus monkey cells to HIV-1, Stremlau et al. showed that human HeLa cells are rendered resistant to infection by transducing them with TRIM5alpha cDNA derived from rhesus monkey cells. Other retroviruses may be restricted by TRIM5alpha in a similar manner.^14,15 Recognition of HIV-1 viral capsid by the B30.2 domain of TRIM5alpha is critical and the site of interaction has been narrowed down to a patch of 13 amino acids which are located in the PRY region of TRIM5alpha.¹⁶ A single amino acid replacement from rhesus to human TRIM5alpha within this region is sufficient to render human TRIM5alpha fully restrictive for HIV-1.^17,18 On the basis of these studies, it is proposed that other members of the TRIM protein family may play a role in innate immunity to retroviruses.^13,14

B30.2: a recent evolutionary adaptation

HMM analyses identify homologous sequences, with the inference that they have a common evolutionary origin, that sequence-similar protein motifs will adopt similar protein conformations and that they are likely to have related functions. The number and diversity of SPRY and B30.2-containing sequences is very large, with no obvious relationship in terms of function. Structural predictions of SPRY and B30.2 domains have been undertaken previously.¹⁹ Both domains are predicted by Seto et al. to adopt a structure resembling the immunoglobulin fold, consisting predominantly of layered antiparallel β sheets.¹⁹ Although this idea is intriguing, it has not been confirmed. SPRY and B30.2 domain sequences can be grouped into 11 protein families encoded by the human genome. We have analysed these protein families in terms of evolutionary conservation in model organisms. SPRY sequences appear to be more ancient, because examples of this domain were identified in the genomes of all eukaryotes, animals, plants and fungi. The B30.2 domain is a fusion of PRY and SPRY domains. It may be regarded as a subset of the group of sequences referred to as SPRY, from which it has evolved and expanded, by incorporating PRY sequences at the amino terminus. The origin of the PRY sequences is not clear from our analysis. The SPRY-PRY combination has occurred relatively recently in evolutionary terms, since the B30.2 domain is only found in vertebrates.

Conclusion

We have analysed all B30.2 and SPRY sequences available in public sequence databases in order to clarify the relationship between these two homologous protein domains. The B30.2 domain has evolved relatively recently from the more ancient SPRY domain by incorporating PRY sequences. This evolutionary adaptation, from SPRY to B30.2 and subsequent expansion in number of genes with this domain, mimics that of some immune receptors, after the emergence of the adaptive immune system.²⁰ The B30.2 domain may have been selected as a component of immune defence and current experimental evidence is consistent with a broad role for the B30.2 domain in innate immune recognition of retroviruses. We conclude that the B30.2 domain is sufficiently distinct from SPRY in terms both of primary structure and emerging function for the separate nomenclature to be maintained.