Distinct evolution process among type I interferon in mammals

Abstract

Interferon (IFN) is thought to play an important role in the vertebrate immune system, but systemic knowledge of IFN evolution has yet to be elucidated. To evaluate the phylogenic distribution and evolutionary history of type I IFNs, 13genomes were searched using BLASTn program, and a phylogenetic tree of vertebrate type I IFNs was constructed. In the present study, an IFNδ-like gene in the human genome was identified, refuting the concept that humans have no IFNδ genes, and other mammalian IFN genes were also identified. In the phylogenetic tree, the mammalian IFN[beta], IFN, and IFN[kappa] formed a clade separate from the other mammalian type I IFNs, while piscine and avian IFNs formed distinct clades. Based on this phylogenetic analysis and the various characteristics of type I IFNs, the evolutionary history of type I IFNs was further evaluated. Our data indicate that an ancestral IFN[alpha]-like gene forms a core from which new IFNs divided during vertebrate evolution. In addition, the data suggest how the other type I IFNs evolved from IFN[alpha] and shaped the complex type I IFN system. The promoters of type I IFNs were conserved among different mammals, as well as their genic regions. However, the intergenic regions of type I IFN clusters were not conserved among different mammals, demonstrating a high selection pressure upon type I IFNs during their evolution.

Protein Cell 2013, 4(5): 383392DOI 10.1007/s13238-013-3021-1 Protein Cell

&

R ESEARCH ARTICLE

Distinct evolution process among type I interferon in mammals

Lei Xu1,2, Limin Yang1, Wenjun Liu1,2,3

1 CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of Microbiology, Chinese Academy of Sciences,

Beijing 100101, China

2 University of Chinese Academy of Sciences, Beijing 100049, China

3 China-Japan Joint Laboratory of Molecular Immunology and Molecular Microbiology, Institute of Microbiology, Chinese Academy

of Sciences, Beijing 100101, China Correspondence: [email protected] (W. Liu), [email protected] (L. Yang)

Received March 12, 2013 Accepted April 8, 2013

Protein Cell

&

ABSTRACT

Interferon (IFN) is thought to play an important role in the vertebrate immune system, but systemic knowledge of IFN evolution has yet to be elucidated. To evaluate the phylogenic distribution and evolutionary history of type I IFNs, 13gen omes were searched using BLASTn program, and a phylogenetic tree of vertebrate type I IFNs was constructed. In the present study, an IFN-like gene in the human genome was identied, refuting the concept that humans have no IFN genes, and other mammalian IFN genes were also identied. In the phylogenetic tree, the mammalian IFN, IFN, and IFN formed a clad e sepa rate f rom the other mammalian type I IFNs, while piscine and avian IFNs formed distinct clades. Based on this phylo-genetic analysis and the various characteristics of type I IFNs, the evolutionary history of type I IFNs was further evaluated. Our data indicate that an ancestral IFN-like gene forms a core from which new IFNs divided during vertebrate evolution. In addition, the data suggest how the other type I IFNs evolved from IFN and shaped the complex type I IFN system. The promoters of type I IFNs were conserved among different mammals, as well as their genic regions. However, the intergenic regions of type I IFN clusters were not conserved among different mammals, demonstrating a high selec tion pressure upon type I IFNs during their evolution.

KEYWORDS type I IFN, evolutionary history, vertebrate, gene cluster

INTRODUCTION

Interferon was rst recognized half a century ago fo r its antiviral

activities (Isaacs and Lindenmann, 1957), a nd its anti-proliferation and immune-regulatory activities were then subsequently discovered (Stark et al., 1998). According to the receptors that they bind, IFNs can be divided into three types: type I, type II, and type III IFNs (Sheppard et al., 2003). There are many different kinds of type I IFNs, such as IFN, , , , , , , and , but type II and type III IFNs are only by a single kind each, IFN and IFN, res pectively (Pestka et al., 2004).

Type I IFNs predominantly function through the typical Jak-Stat pathway. When type I IFNs bind to their high-afnity receptors, a heterotrimer named interferon-stimulated gene factor 3 (ISGF3) is formed to activate the expression of interferon-stimulated gene (ISGs). Aside from the canonical Jak-Stat pathway, IFNs can also activate the MAPK pathway (Stancato et al., 1997; David, 2002). However, due to its complex activities in the immune system, new IFN pathways an d roles remain to be discovered.

Type I IFNs have been identified in zebrafish, Atlantic salmon, grass carp, and Fugurubripes, which several of them display similar activities to their mammal homologs (Altmann et al., 2003; Robertsen et al., 2003). IFN is hypothesized to have arisen from birds, and the other type I IFNs are present only in mammals (Sick et al., 1996).IFN, also known as limitin, was discovered for its abi lity to arrest the growth of or kill lympho-hematopo ietic cells, and is thought to only exist in mice (Oritani et al., 2000). IFN has been described in sheep, pigs, and horses and is expressed at 15 days of gestation in conceptus (Lefvre and Boulay, 1993; Cochet et al., 2009). IFN is considered to only exi st in ruminants but is present in the human genome as a pseudogene (Whaley et al., 1994). Amphibian type I IFN genes were inferred as intron-containing and 4-cysteine-containing IFN genes, s imilar to zebrash IFNb and IFNc, while reptile type I IFNs were inferred from genomic se-

Higher Education Press and Springer-Verlag Berlin Heidelberg 2013 May 2013 | Volume 4 | Issue 5 | 383

RESEARCH ARTICLE

Lei Xu et al.

quence as intronless forms (Sun et al., 2009; Qi et al., 2010).

Robert has hypothesized a good model for type I IFN evolution, which suggests that IFN diverged from IFN and that IFN arose from IFN approximately 36 million years ago (Roberts et al., 1998). The evolution of IFNis well described, and both gene conversion and duplication played important roles in the formation of the IFN gen e family (Woelk et al., 2007). However, the detailed phylogenic distribution and evolutionary history of type I IFNs in vertebrates remains unclear (Qi et al., 2010).Towards a better understanding of type I IFN evolution, we repor t several novel IFNs and illustrated the evolution histo ry of type I IFNs in vertebrates. The gene synteny analysis demonstrates a high selection pressure upon type I IFNs during their evolution.

RESULTS

Type I IFN gene clusters in genomes

Different types IFN genes are located on di ff erent chromosomes. In humans, type I IFN genes represent a gene cluster on chromosome 9 that includes 14 IFN genes, one IFN gene, one putative IFN gene, one IFN pseudogene, one IFN gene and one IFN gene. The I FN gene is not adjacent to this gene cluster, and a similar situation was foun d in other mammals. In mouse, type I IFNs form a gene cluster on chromosome 4, which includes 16 IFN genes, one IFN pseudogene, 14 limitingenes, and one IFN gene. In most species, the IFN and IFN genes are located on the two extremities of type I IFN clusters, while the other type I IFN g e nes were randomly located relative to one another, except limitin, which forms a small cluster within the type I IFN cluster.

As indicated by the B LAST results, IFN and IFN genes may not exist in the mouse and dog genomes and this result in not due to lack of sequence coverage as the mouse and dog genomes have been fully characterized (Fig. 1), which is consistent with previous studies (Hardy et al., 2004). Limitins only exist in mice. Indeed, even in the rat genome, the limitin homolog is a pseudogene with an early stop codon and considerably degraded sequence (date not shown).

A putative IFN gene, containing a pseudogene named IFN12p, was identied in the human genome (Fig. 2). This gene is loca ted in the type I IFN gene cluster and represents a single gene that is quite different from the other IFN genes in sheep and pigs. Sharin g 62.3% similarity with horse IFN1 and <50% similarity with the other human IFNs, the putative human IFN gene is likely a new subtype of human type I IFN (Table 1). Located immediately beside the IFN gene lies the IFN8 gene and IFN11p, which is unexpressed in humans (Henco et al., 1985). In addition, the homolog of human pseudogene in other primates is similar with itself, represents as a single gene with an early stop codon (date not shown).

Four IFN genes were identied in the cow, horse, cat, and dog genomes (Fig. 2A ) . All of these genes display >75% similarity with human IFN and <50% similarity with the other IFNs previously identied in these genomes. Similar to human IFN,

these four IFN genes are located on the extremities of the type I IFN clusters, i.e., there is high conservation of the IFN genomi c location. In addition, three IFN genes were identied in the horse, cat, and dog genomes (Fig. 2B). Both the phylo-genetic tree and sequence similarity indicate that these genes are true IFN genes. All of these genes are located outside o f the type I IFN gene cluster, just like the human IFN gene.

The complex relationship among type I IFNs in vertebrates

Within the type I IFN grouping, avian type I IFNs forms a separate clade from piscine and mamm alian type I IFNs (Fig. 3). In our analysis, piscine type I IFNs are the outgroup of avian and mammalian type I IFNs, consistent with the course of vertebrate evolution. Unique to piscine IFNs, the divergence pattern of type I IFNs does not match the speciation of sh, and this

Protein Cell

&

Type I IFN

Human

Cattle

Sheep

Pig

Horse

Cat

Dog

Mouse

Chicken

Fish

Present

Putative

Pseudogene Absent

Figure 1. Absence/presence plots in a subset of 10 vertebrate genomes. Type II IFN and IFN exist in all vertebrates, while type III IFN is absent in sh genomes. IFN can be found in both avian and mammalian genomes, and the other type I IFN only exist in mammal. Mouse and dog lost their IFN genes and IFN genes during species evolution, but limitins arise in mouse genome. 4 IFN genes and IFN genes was inferred, suggesting that IFN and IFN exist in all mammals. IFN only exist in ruminants.

384 | May 2013 | Volume 4 | Issue 5 Higher Education Press and Springer-Verlag Berlin Heidelberg 2013

Distinct evolution process among type I interferon in mammals

RESEARCH ARTICLE

A

B

C

: 113 : 113 : 113 : 112 : 112

Cow IFNE: Human IFNE:

Horse IFNE:

Dog IFNE:

Cat IFNE:

Cow IFNE: Human IFNE:

Horse IFNE:

Dog IFNE:

Cat IFNE:

* 20 * 40 * 60 * 80 * 100 *

120 * 140 * 160 * 180 * 200

: 193 : 208 : 193 : 187 : 208

Dog IFNK:

Cat IFNK:

Human IFNK:

Horse IFNK:

* 20 * 40 * 60 * 80 * 100 *

120 * 140 * 160 * 180 * 200

Horse IFNK:

Human IFNK:

Dog IFNK:

Cat IFNK: : 208

: 208

: 113

: 113

: 113

: 112

: 209

: 208

: 207

Cow IFNDp :

Cat IFNDp :

Human IFND :

Pig IFND1 :

* 20 * 40 * 60 * 80 * 100

* 120 * 140 * 160 * 180 * 200 *

* 440 * 460 * 480 * 500 * 520 *

: 75: 105 : 87: 102

: 167 : 197

: 197

: 188

: 273

: 303

: 294

: 302

: 352 : 393 : 400 : 388

: 453 : 494 : 501 : 490

Pig IFND1 :

Human IFND :

Cow IFNDp :

Cat IFNDp :

Human IFND :

Pig IFND1 :

Cow IFNDp :

Cat IFNDp :

Human IFND :

Pig IFND1 :

220 * 240 * 260 * 280 * 300 3

20 * 340 * 360 * 380 * 400 * 420

Protein Cell

&

Cow IFNDp :

Cat IFNDp :

Human IFND :

Pig IFND1 :

Cow IFNDp :

Cat IFNDp :

540 * 560 * 580 *

Pig IFND1 :

Human IFND :

Cow IFNDp :

Cat IFNDp :

: 471 : 513 : 562 : 509

Figure 2. Multiple alignment and gene synteny analysis of novel identied IFNs from human, cat, horse, cow and pig. The alignment was generated by CLUSTALW and then edited by GENEDOC. (A and B) Multiple alignment of novel identied IFN and IFNs. The putative signal peptide sequences predicted by SignalP4.0 are underlined and conserved cysteines are marked with black trilateral. Conserved regions are shadowed. (C) Multiple alignment of novel identied IFN pseudogene.

disparity was not resolved by restricting the list of piscine type I IFN species in the phylogenetic analyses. Thus, the rea son for this discrepancy is the dive rgence of 2C-containing IFNs and 4C-containing IFNs in sh.

The avian and mammalian IFNs form distinct species-specic clades in the phylogenic tree (as the piscine IFNs do). Even though IFNs have been puried from reptiles and have similar biophysical properties as mammalian IFNs, the failure to clone these IFNs makes it difficult to further study reptile type I IFNs.

The major subgroups of IFNs identied thus far form subgroups within the mammalian type I IFN family. Phylogenetically, the mammalian type I IFN subtypes form clades consistent with mammalian speciation in the phylogenetic tree. The rst diverging group of IFNs within the mammalian type I IFN clade is that of the unduplicated IFNs. IFN forms an outgroup in this subclade, which indicates a different evolutionary route compared to IFN and IFN. The next subgroup to diverge from the remaining mammalian type I IFNs is that of IFN and IFN. These two subgroups may in fact be more related to

each other tha n to other mammalian type I IFN subtypes.

Both IFN and IFN form an outgroup from the remaining mammalian IFNs. Like the pis cine IFNs, the dive rgence pattern o f porcine IFN genes does n ot form a single clade in the phylogenetic tree, suggesting a close evolutionary distance among IFN genes from different species. The failure to nd lim itin in the pig and horse genomes, which have several IFN gene s, and the failure to nd IFN genes in mice and rats, sugge sts a preferential evolutionary relationship between IFN and IFN.

IFN and IFN form a subgroup in the phylogenetic tree, and porcine and bovine IFN, together with ovine and bovin e IFN, fo rm a core clade in this subgroup. Though IFN and IFN display intimate relationships, their functions are quite different: IFN is an antiviral and immuno-regulator, just like IFN, while the preferential function of IFN is to ensure the pregnancy continues through preventing the corpus luteum from degradation. Aside from their differences in function, the species dist ribution o f IFN is more extensive than that of IFN, which is only found in sheep and cattle. According to the phylogenetic tree and the characteristics of IFN and IFN,

Higher Education Press and Springer-Verlag Berlin Heidelberg 2013 May 2013 | Volume 4 | Issue 5 | 385

RESEARCH ARTICLE

Lei Xu et al.

Table 1. Sequence identities and divergences among type I IFNs

CalfA1 EqcA1 FecA1 HosA1 BotB CalfB FecB EqcB HosB EqcD1 HosE

CalfA1 - 69.5 73.2 65.2 32.3 35.8 39 38.7 32.8 44 36.2

EqcA1 30.3 - 73.5 79.8 34.2 37.3 39.1 40.7 39.6 51.8 40

FecA1 25.4 22.9 - 69.8 35.1 36.2 37.8 38.5 34.4 46.6 36.8

HosA1 36.4 19.6 27.3 - 36.4 36.4 39.2 40.5 36.5 52.6 42.5

BotB 93.6 88.1 87.8 91.1 - 65.8 67 70.1 65.4 33.7 39.6

CalfB 85.6 85 88.4 88.6 38.6 - 82 74.2 71.5 36.7 41

FecB 79.2 77.9 79.4 81.5 35.4 17.5 - 75.4 71.7 34.5 41.2

EqcB 77.8 73.5 76.2 77.9 31.8 28.4 25.2 - 73.8 37.3 40.5

HosB 87.3 80.6 90 86.5 36.8 30.2 28.7 27.6 - 34.7 43.1

EqcD1 65.8 56.1 62.6 55.7 92.1 86.7 94.1 84.5 86.1 - 38.3

HosE 90.5 81 88 73.8 87.3 81.1 82.8 76.1 77.3 90.1 -

EqcW1 51.4 41.6 47.5 41.8 83.7 78.4 74.3 68.4 69.1 58.1 75.2

FecW1 53.1 47.7 51.6 50.3 91.3 89.3 83.4 79.6 87.6 65.6 93.2

HosW 51.2 39.1 42.3 38.9 79.3 74 72.2 64.4 73.7 52.4 77.8

HosT 45.8 36.9 39.4 37.5 83.4 77.3 75.6 66.6 76.6 59 75

FecDp 86 71.1 83.2 71.9 87.8 82 91.1 85 98.4 30.6 82.9

HosD 79.7 71.3 75.5 63.2 89.5 86.9 96.4 90 94.4 28.9 96.1

CalfE 91.3 76.4 89.2 77.5 92.2 88.1 85.1 83.6 80.7 88.6 19.6

CalfK 126.8 111.3 122.1 101.8 144.3 124.2 127.9 136.8 123.1 129.8 108.1

EqcE 96.7 82.8 91.1 79.3 84.3 79.1 81.4 75.2 78.3 91 16.7

EqcK 122.6 110.1 117.6 105.3 122 116.5 112.1 119.8 114.4 132.7 107.2

FecE 92.5 81.2 91.2 81.1 91.2 83.3 81.9 80.2 77.4 89.7 22.1

EqcW1 FecW1 HosW HosT FecDp HosD CalfE CalfK EqcE EqcK FecE

CalfA1 54.1 51.2 53.7 58 34.6 32.4 35.3 35.3 34.6 35.5 36.5

EqcA1 62.3 56.9 63.2 64.3 41.6 40 43.2 41.8 40.9 43.1 43.4

FecA1 54.9 51.9 59.3 60 34.3 35.4 37.2 36.8 37 35.4 39.3

HosA1 61.8 54.9 63.5 65.3 38.6 45 42.9 37.9 42.6 39.5 44.2

BotB 37.3 31.6 38.5 39.2 31.6 31.9 35.5 28.9 39.2 32.6 36.5

CalfB 38.7 35.5 40.8 39.6 33.1 33.3 38.9 33.3 39.9 35.7 39.8

FecB 41.5 37.8 41.7 41.4 30.7 31 38.9 36.5 40.3 38 40.8

EqcB 39.9 38.3 43.5 43.1 28.9 32.8 36.9 36.5 40.6 37.1 38.3

HosB 43.3 33.7 40.1 41 30.4 28.8 39.9 36.9 41.5 39.9 40.2

EqcD1 50.2 44.8 51.6 48.2 62.3 64.3 39.3 34.1 39.1 34.1 40.1

HosE 41.2 37.1 40.3 40.6 36.7 36.1 79.4 36.2 83.5 34.6 76.9

EqcW1 - 61.7 69 69.6 41 41.3 41.8 36.6 41.4 37.6 40.8

FecW1 40.1 - 66.7 70.4 36.7 35.8 38.7 33.3 37.8 31.6 37.7

HosW 33.6 34.5 - 73.8 44 43.8 41.5 39.8 40.2 39.6 40

HosT 31.7 30.1 27.3 - 44.6 38.4 43.3 37.9 41.8 38.4 44.4

FecDp 66 73.3 63.7 63 - 49.7 38.3 28 36.4 28.3 38.3

HosD 70.1 80.2 60.9 72.6 36.3 - 31.5 31.7 31.9 28.1 32.9

CalfE 78.8 96.1 85.3 77.9 84.2 103.9 - 39 82.6 36.7 88.3

CalfK 128.1 121.3 103 121.3 135.1 142.4 99.4 - 37.5 82.1 32.2

EqcE 74.7 93.3 82.2 77.8 84 98.6 15.4 100 - 37.1 80.9

EqcK 136.6 130 103.2 117.4 127 145.3 98.9 16.8 99.5 - 32.9

FecE 81.6 96.7 92.5 76.4 80.8 108.6 11.8 108.5 17.3 101.9 -

Percent similarity in upper triangle and percent divergence in lower triangle

386 | May 2013 | Volume 4 | Issue 5 Higher Education Press and Springer-Verlag Berlin Heidelberg 2013

Protein Cell

&

Distinct evolution process among type I interferon in mammals

RESEARCH ARTICLE

A B

C

D

EqcD2

EqcD1

HosD2

FecDp

PtvD

OvaD BotD

SusD3 SusD10 SusD4

SusD1 SusD2

SusD7 SusD8

SusD6 SusD11

91 61 99 SusD5 86 SusD9

100

54

13 13

45

38

34

100

IFN

45

100

100

CrgE MumE FecE

CaltE

HosE EqcE

BotE SusE

Protein Cell

100

90 MamK

100

IFN and

Mammal IFN

&

64

58

60

30

63

IFN and

97

100

MumK RanK

CrgK

IFN and

14

32 16

21

10

11

BosK

SusK

EqcK FecK PtvK

HosK

CaltK

IFN

100

Bird IFN

Fish IFN

Figure 3. Phylogenetic tree analysis of all type I IFNs including the novel identied IFNs. (A) Overview for the phylogenic tree of type I IFNs. Symbols inside the IFN clade are human, bat, cow, pig, cat, dog, mouse, respectively. (B) Details for IFN clades. (C) Details for the IFN clades. (D) Details for the IFN clades. Bootstrap values over 75% are shown.

IFN maybe the last mammalian type I IFN to evolve.

IFN generally forms species-specific clades in the IFN subgroup unless the two species in question are closely related. This renders direct homologs of IFN between different species difcult to nd, suggesting a high selection pressure on IFN during evolution.

Overview of type I IFN evolution

On the basis of the phylogenetic analysis above and the known characteristics of type I IFNs, we propose a hypothesis for type I IFN evolution: The ancestral type I IFN gene likely

contained introns, similar to piscine type I IFNs, and was liable to duplicate which seems to be the ancestor of IFN genes instead of IFN genes. Synchronously with species evolution, type I IFN genes retain original characteristics in amphibian genomes. IFN evolved from the ancestral IFN-like genes soon after reptiles arose (Fig. 4). This ancient IFN gene may be encoded by several exons, as no obviously homologs of avian IFN genes can be found in the green lizard genome, and this gene represents the ancestor of avian and mammalian IFN genes, as avian type I IFNs are homolog of mammalian type I IFNs (Sick et al., 1998). Consistent with species evolution from

Higher Education Press and Springer-Verlag Berlin Heidelberg 2013 May 2013 | Volume 4 | Issue 5 | 387

RESEARCH ARTICLE

Lei Xu et al.

Fish IFN

Amphibian IFN

Reptile IFN

Reptile IFN

Bird IFN

Mammal IFN

Mammal IFN Mammal IFN

Bird IFN

Mammal IFN

Mammal IFN

Mammal IFN

Mammal IFN

Mammal IFN

Figure 4. Evolution history of type I IFNs. The type IIFN genes of sh and Amphibian contain 45 introns. No reptile IFN genes were cloned as recent studies, but it is no doubt that the IFN genes do exist as the completeness of the revolution routes. The red circle donates a gene with introns, and dotted line represents a gene or a route which is inferred.

Protein Cell

&

reptiles to birds and mammals, a retrotr ansposition occurred in the type I IFNs other than IFN, as evident by the fact that most bird and mammal type I IFNs are intronless. IFN genes may have arisen from a putative reptile IFN gene before this retrotransposition occurred, and avian type I IFNs remain in their ancestral state. In contrast, mammals encountered more evolutionary pressure to survive in tougher environments, and their type I IFNs divided into many subtypes.

IFN genes duplicated, and one of them represented the ancestor of mammalian IFN. This subtype of type I IFNs likely arose at the very beginning of mammalian e volution as practically every mammal possess IFN and IFN. Afterward, IFN, tog ether with IFN and IFN, remained untouched as duplications of these genes seem to be difcult (only platypus and cow have two IFN-like genes) for unknown reasons.

As avian IFN and mammalian IFN genes are strict homo-logues of the ancient IFN-like genes and IFN genes show broader distribution than the other multigene Type I IFNs, the complex mammalian Type I IFN system should be shaped by ancient mammalian IFN genes. In addition, mammalian IFN exhibits common activities compared with the other mammalian IFNs. In contrast to IFN, it is much easier to duplicate IFN as approximately 10 IFN genes are present in most vertebrates. With selective pressure, some of these genes gained specic functions. One of these genes became the foreru nner of IFN and duplicated into a cluster of genes. Afterward, a few of the IFN genes in the ruminant ancestor diverged into IFN to help form placentas. One of the ancient IFN genes evolved into the forerunner of the IFN and IFN genes. The divergence of IFN and IFN is lik ely different results of the same selection pressure on different species. During this selection, IFN gained a reproductive resp onse element in the porcine and ovine genomes, while it failed in the mouse genome.

Synteny alignment of the mouse, dog, pig, and cow type I IFN clusters against the human type I IFN cluster

The type I IFN cluster is not strictly conserved in all mammal genomes, especially itsintergenic regions. Part of the limitin genic regions displays limited similarity with human IFN, and all of the limitin genes are located together in the type I IFN cluster (Fig. 5).This indicates that the master force for limitin gene family conformation was duplication instead of conversion, and this kind of duplication occurred after the speciation of mice. All mouse IFN genes have homologous regions in the human genome, and most of the upstream region of these genes is conserved, ensuring that IFN can be induced by viral infection and pathological processes. The corresponding region of human IFN in the mouse genome seems to be mouse IFN14, which is evidence for the hypothesis that IFN evolved from IFN.

IFN cannot be identied in the human type I IFN cluster, and only a previously described pseudogeneis located outside of the gene cluster on the IFN side. However, a discontinuous region of the human genome displays limited conservation with IFN genes, and further analysis revealed that these regions are microRNAs, such as the MIR31 host gene.

DISCUSSION

In the present study, an IFN-like gene in the human genome was identied, refuting the concept that humans have no IFN genes (Pestka et al., 2004). This gene is located on chromo-some 9, with a pseudogene named IFNA12p inside of it. Both phylogenic and sequence similarity analyses demonstrated that this gene is an IFN gene. In the phylogenic tree, the mammalian IFN, IFN, and IFN genes formed a clade separate from the other mammalian type I IFNs, while piscine and avian IFNs formed distinct clades. Based on our phylogenetic

388 | May 2013 | Volume 4 | Issue 5 Higher Education Press and Springer-Verlag Berlin Heidelberg 2013

Distinct evolution process among type I interferon in mammals

RESEARCH ARTICLE

A B C

D

IFNE

IFNA1

IFNA5

IFNE11

Gm13287

Gm13277

Gm13290

Gm13286

Gm13282

IFNAb

Gm13280

Gm13281

IFNA12

IFNA14

IFNB

405209

Gm12597

IFNE

IFNA7

LOC611406

LOC100688084

LOC100687928

LOC100687692

LOC100686162

IFNB

LOC100737367

LOC100157354

LOC100156936

LOC100622159

LOC100154075

LOC100739443

LOC100154912

LOC100155323

LOC100624378

LOC100736739

LOC100736907

405209

212046

IFNE

IFNW3

IFNA1

IFND5

IFNA4

IFND2

IFND9

IFND4

IFNAW

405209

1

1

1

1

1048877

1

358177

Pig

1

Moues

Dog

LOC100847720

LOC618801

LOC513706

LOC616272

LOC618947

LOC781853

LOC617135

LOC618985

LOC100298530

LOC100299481

LOC783912

LOC616977

LOC787343

Figure 5. Dot-plot analysis reveals a complex pathway of type I IFN evolution. Dotplot analysis was applied to the genic and intergenic regions of the mouse (A), dog (B), pig (C) and cattle (D) type I IFN gene clusters to human type I IFN cluster. The type I IFN genes are cited as NCBI shows. IFN-like proteins are also cited. The alignment shows as intermittent lines which represent the regions with at least 70% sequence identity, indicating that the type I IFN cluster is not strictly conserved in all mammal genomes, especially its intergenic regions.

Protein Cell

IFNE

IFNT

IFNT2

IFNT

IFNAG

IFNAH

IFNB1

405209

669175

&

1

1

Cattle

analysis and the known characteristics of the various type I IFNs, we hypothesized the evolutionary history of type I IFNs. In this evolution route, IFN formed a core from which new IFNs divided during vertebrate evolution. One of the IFN genes evolved away from the other genes and became the ancestor of IFN in reptiles, and most of these genes lost introns through retrotranspositions when birds and mammals diverged from reptiles. The ancestor IFN-like genes duplicated more frequently than the ancestor IFN-like gene and divided into several subtypes (including IFN, IFN, and IFN) during the evolution of some animals, while IFNevolved into IFN and I FN at the very beginning of the emergence of mammals.

The doubts and suspicions in this evolutionary model focus on the time when the type I IFN introns disappeared and whether reptiles genomes encode IFN gen es. Type I IFNs exist in all kinds of vertebrates, but only the sh and amphibian type I IFN genes contain four introns as recent research (Qi et al., 2010). In the sh genome, IFNa, IFNb, IFNc and IFNd are considered to be different kinds of type I IFNs that only exists in sh and not strictly orthology of higher vertebrates IFN genes. Both IFNa, IFNb, IFNc and IFNd displays a considerable distance to the other type I IFNs in our phylogenetic tree, but IFNa and IFNc show similar antiviral activities and ability to induce antiviral genes, like mammalian IFN do, while IFNb and IFNd show little activities (Sun et al., 2009; Svingerud et al., 2012).

When the introns of these type I IFN genes disappeared remains unknown, but the IFN genes in amphibians may have

introns (Qi et al., 2010). It is very likely that reptile genomes encode IFNs for the sake of the completeness of evolution routes, even though no reptiles IFN gene has yet been cloned, only been inferred. The failure to detect reptile IFNs when using chicken IFN genes with BLAST against the green lizard genome suggests that IFN genes in reptiles are much similar to fish IFN genes than those in warm-blooded animals. However, the IFN genes of mammals have one intron each, suggesting that the existence of intron-containing reptile IFN genes exist.

Amphibian IFNs represent as intron-containing IFNs with ve different molecules as inferred by present research, however, functional study on frog type I IFNs is still needed (Qi et al., 2010). A retrotransposition event likely occurred to type I IFNs during reptile evolution as avian and mammalian type I IFNs are intronless genes. Type I IFN genes in lizards may contain introns as no obvious homologs of avian and mammalian type I IFN genes can be found in the green lizard genome. The situation in snakes and turtles remains unknown, and perhaps intron-containing and intronless type I IFN genes coexist in their genomes. IFN genes likely exist in reptile genomes because these genes appear to exist in all avian and mammalian genomes. The IFN gene may have duplicated during reptile evolution into mammals, and retrotransposition occurred simultaneously. Further, the IFN genes in mammals each contain an intron, and IFN genes are intronless, suggesting the possibility that duplicated IFN genes or the ancestor of

Higher Education Press and Springer-Verlag Berlin Heidelberg 2013 May 2013 | Volume 4 | Issue 5 | 389

RESEARCH ARTICLE

Lei Xu et al.

IFN genes exist in reptilian genomes.

Even though all type I IFNs bind to the same receptors, different IFN display different bioactivities. The functions of type I IFN maintain good relation with their evolutions. The antiviral activity and anti-proliferation activity of IFN is lower than most of the IFNs, and the antiviral activity of IFN-evoluted IFNs is stronger than that of IFN and IFNs (Sang et al., 2010; Lavoie et al., 2011). The ISGs that IFN induces tend to regulate host immunoreactions, while IFN tends to induce antiviral proteins, such as 2-5 OAS and PKR (Qu et al., 2013). This may be caused by the different ways that they interact with IFN receptor subunits. Despite their differences in sequence, IFNdemonstrates little difference with IFN in their expression, function, ternary complex structures with receptors, and abilities to induce ISGs (Thomas et al., 2011). Expressed by trophoblasts before they attach to the placenta, IFN is important to prevent the corpus luteum from degradation and, hence, ensuring the pregnancy continues, while IFN is expressed in conceptus and shows a potent ability to regulate pregnancy (Chelmoskasoyta, 2002; Roberts, 2007).It has been hypothesized that IFN is not virally inducible a nd that its function has no relationship to pathogenesis, which is quite different from the other multigene Type I IFNs, but recent research shows that bovine IFNis an antiviral protein capable of inducing 2-5 OAS with less toxicity to the cells, which suggests that IFN could be a better drug than IFN for patients suffering from viruses(Johnson et al., 2001). The IFN-evoluted IFNs, IFN and IFN, tends to induce effectors which can regulate host immunoreactions. Over expression of IFN in pancreatic islets can induce diabetes in mouse, and recent research has shown that the expression of IFN in skin has a close association with systemic lupus erythematosus and inflammation (Vassileva et al., 2003; Harley et al., 2010). IFN can enhance the lymph ocyte recruitment to lung alveoli with reduced inammation, promote migration of antigen-specic CD8+ T cells to the gut, which suggests IFN an important role in mucosal immunity (Xi et al., 2012). This relationship between evolution and function of type I IFN ma y provide a new view for type I IFN function analysis.

MATERIALS AND METHODS

Genome selection

A total of 496 IFN sequences from 99 species were acquired from NCBI(http://www.ncbi.nlm.nih.gov/). Allelic genes and pseudogenes were omitted for the sake of phylogenetic analyses. The IFN genes from human, mouse, and pig have already been characterized in detail before, and some of the features of IFN genes from ruminants and IFN genes from pigs have also already been described (Bazer et al., 1997). To nd undetected IFN genes, BLASTn (Buhler et al., 2007) was used to search 13 genomes (human, mouse, dog, cat, sheep, pig, cow, horse, chicken, zebrash, green anole, Nile tilapia, and western clawed frog). Human IFN2b (AY255838.1), IFN (X58822.1), IFN (M28622.1), IFN (AY190045.1), and IFN (AF315688.1), mouse IFN1 (NM_197889.2), pig IFN1 (GQ415074.1), and cow

IFN1 (AF238611.1), chicken IFN1 (AB021153.1) and chicken IFN (AY974089.1) were chosen as initial queries. All of the default alignment/search parameters were used. Sequences displaying >10% similarity were collected for further identication. Short homologous sequences were extended to 600 bp to calculate the similarity with given sequences.

Phylogenetic analyses

All of the IFN sequences were aligned with ClustalW (EBI, www.ebi. ac.uk/clustalw) (Thompson et al., 2002). After manual correction of the sequences, a neighbor-joiningtree was constructed using MAGE 5.0 with the following parameters:method = NJ, substitution model = Poisson correction method, and 1000 bootstrap replicates (Tamura et al., 2011). We also constructed a maximum likelihood tree with the following parameters to evaluate the reproducibility of the grouping of the IFN genes:method JTT and 1000 bootstrap replicates.

Gene synteny analysis

MultiPipmaker was used to align both the genic and intergenic regions of the mouse, dog, pig, and cow type I IFN gene clusters to the human type I IFN cluster (Schwartz et al., 2003). Default parameters were used except that the search one strand and single coverage options were chosen. Genes were cited as shown by NCBI.

Competing Interests: The authors have declared that no competing interests exist.

ACKNOWLEDGEMENTS

This study was funded by the National Natural Science Foundation of China (Grant No. 31100644), the Ministry of Science and Technology program of China (Grant Nos. 2011AA10A215, and 2010GB24910698). Wenjun Liu is a principal investigator of the National Natural Science Foundation of China Innovative Research Group (Grant No. 81021003). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

ABBREVIATIONS

Bot, Bostaurus Cow; Crg, Cricetulusgriseus Hamster; Calf, Canis lupus familiaris Dog; Eqc, Equuscaballus Horse; Fec, Feliscatus Cat; Hos, Homo sapiens Human; ISG, interferon-stimulated gene; ISGF, interferon-stimulated gene factor; Mum, Musmusculus Mouse; Ova, Ovisaries Sheep; Ptv, Pteropus vampyrus Bat; Ran, Rattusnorvegicus Rat; Sus, Susscrofa Pig

COMPLIANCE WITH ETHICS GUIDELINES

Xu Lei, Yang Limin, and Liu Wenjun declare that they have no conict of interest.

This article does not contain any studies with human or animal subjects performed by the any of the authors.

REFERENCES

Altmann, S.M., Mello n, M.T., Distel, D.L., and Kim, C.H. (2003). Molecular and functional analysis of an interferon gene from the zebrash, Danio rerio. J Virol 77, 19922002.

Protein Cell

&

390 | May 2013 | Volume 4 | Issue 5 Higher Education Press and Springer-Verlag Berlin Heidelberg 2013

Distinct evolution process among type I interferon in mammals

RESEARCH ARTICLE

Bazer, F.W., Spencer, T.E., and Ott, T.L. (1997). Interferon tau: A novel pregnancy eecognition signal. Am J Reprod Immunol 37, 412420.

Buhler, J.D., Lancaster, J.M., Jacob, A.C., and Chamberlain, R.D.

(2007). Mercury BLASTN: Faster DNA sequence comparison using a streaming hardware architecture. In Recongurable Systems Summer Institute, Oxford University Press, Oxford, 18. Chelmoskasoyta, A. (20 02). Interferon tau and its immunobiological role in ruminant reproduction. Arch Immunol Ther Exp 50, 4752. Cochet, M., Vaiman, D., and Lefvre, F. (2009). Novel interferon delta genes in mammals: Cloning of one gene from the sheep, two genes expressed by the horse conceptus and discovery of related sequences in several taxa by genomic database screening. Gene 433, 8899.

David, M. (2002). Signal tr ansduction by type I interferons. Biotechniques (Suppl), 5865.

Hardy, M.P., Owczarek, C.M., Jermiin, L.S., Ejdebck, M., and Hertzog, P.J. (2004). Characterization of the type I interferon locus and identication of novel genes. Genomics 84, 331345.

Harley, I.T.W., Niewold, T.B., Stormont, R.M., Kaufman, K.M., Glenn, S.B., Franek, B.S., Kelly, J.A., Johnson, G.A., Stewart, M.D., Gray, C.A., Choi, Y., Burghardt, R.C., Yu-Lee, L.Y., Bazer, F.W., and Spencer, T.E. (2001). Effects of the estrous cycle, pregnancy, and interferon tau on 2,5-oligoadenylate synthetase expression in the ovine uterus. Biol Repr 64, 13921399.

Henco, K., Brosius, J., Fujisawa, A., Fujisawa, J.I., Haynes, J.R., Hochstadt, J., Kovacic, T., Pasek, M., Schambock, A., Schmid, J., et al. (1985). Structural relationship of human interferon alpha genes and pseudogenes. J Mol Biol 185, 227260.

Kilpatrick, J.R., Hutchings, D., Divers, J., et al. (2010). The role of genetic variation near interferon-kappa in systemic lupus erythematosus. J Biomed Biotechnol 2010, 111.

Isaacs, A., and Lindenmann, J. (195 7). Virus interference. I. The inter-feron. Proc Roy Soc B: Biol Sci 147, 258267.

Lavoie, T.B., Kalie, E., Crisafulli- Cabatu, S., Abramovich, R., DiGioia, G.,

Moolchan, K., Pestka, S., and Schreiber, G. (2011). Binding and activity of all human alpha interferon subtypes. Cytokine 56, 282289. Lefvre, F., and Boulay, V. (1993). A novel and atypical type one inter-feron gene expressed by trophoblast during early pregnancy. J Biol Chem 268, 1976019768.

Oritani, K., Medina, K.L., Tomiyama, Y. , Ishikawa, J., Okajima, Y.,

Ogawa, M., Yokota, T., Aoyama, K., Takahashi, I., Kincade, P.W., et al. (2000). Limitin: An interferon-like cytokine that preferentially inuences B-lymphocyte precursors. Nat Med 6, 659666. Pestka, S., Krause, C.D., and Walter, M. R. (2004). Interferons, interferon-like cytokines, and their receptors. Immunol Rev 202, 832.

Qi, Z.T., Nie, P., Secombes, C.J., and Zou, J. (2010). Intron-containing type I and type III IFN coexist in amphibians: Refuting the concept that a retroposition event gave rise to type I IFNs. J Immunol 184, 50385046.

Qu, H., Yang, L., Meng, S., Xu, L., Bi, Y., Jia, X., Li, J., Sun, L., and Liu, W. (2013). The differential antiviral activities of chicken interferon (ChIFN-) and ChIFN- are related to distinct interferon-stimulated gene expression. PLoS ONE 8, e59307.

Roberts, R.M., Liu, L., Guo, Q.T., Leaman, D., and Bixby, J. (1998). The evolution of the type I interferons. J Interferon Cytokine Res 18, 805816.

Roberts, R.M. (2007). Interferon-tau, a type 1 i nterferon involved in maternal recognition of pregnancy. Cytokine Growth Factor Rev 18, 403408.

Robertsen, B., Bergan, V., Rkenes, T., Larsen, R ., and Albuquerque, A.

(2003). Atlantic salmon interferon genes: cloning, sequence analysis, expression, and biological activity. J Interferon Cytokine Res 23, 601612.

Sang, Y., Rowland, R.R., Hesse, R.A., and Blecha, F. (2010). Differential expression and activity of the porcine type I interferon family. Physiol Genomics 42, 248258.

Schwartz, S., Elnitski, L., Li, M., Weirauch, M., Ri emer, C., Smit, A.,

Program, N.C.S., Green, E.D., Hardison, R.C., and Miller, W. (2003). MultiPipMaker and supporting tools: alignments and analysis of multiple genomic DNA sequences. Nucleic Acids Res 31, 35183524. Sheppard, P., Kindsvogel, W., Xu, W., Henderson, K., Schlutsmeyer, S.,

Whitmore, T.E., Kuestner, R., Garrigues, U., Birks, C., Roraback, J., et al. (2003). IL-28, IL-29 and their class II cytokine receptor IL-28R. Nat Immunol 4, 6368.

Sick, C., Schultz, U., Mnster, U., Meier, J., Kaspers , B., and Staeheli, P.

(1998). Promoter structures and differential responses to viral and nonviral inducers of chicken type I interferon genes. J Biol Chem 273, 97499754.

Sick, C., Schultz, U., and Staeheli, P. (1996). A family of genes coding for two serologically distinct chicken interferons. J Biol Chem 271, 76357639.

Stancato, L.F., Sakatsume, M., David, M., Dent, P., Dong, F., Petricoin,

E.F., Krolewski, J.J., Silvennoinen, O., Saharinen, P., et al. (1997). Beta interferon and oncostatin M activate Raf-1 and mitogen-activated protein kinase through a JAK1-dependent pathway. Mol Cell Biol 17, 38333840.

Stark, G.R., Kerr, I.M., Williams, B.R.G., Silverman, R.H. , and Schreiber, R.D. (1998). How cells respond to interferons. Annu Rev Biochem 67, 227264.

Sun, B.J., Robertsen, B., Wang, Z.Q., and Liu, B. (2009). I dentication of an Atlantic salmon IFN multigene cluster encoding three IFN subtypes with very different expression properties. Dev Comp Immunol 33, 547558.

Svingerud, T., Solstad, T., Sun, B.J., Nyrud, M.L.J., Kileng , ., Greiner-

Tollersrud, L., and Robertsen, B. (2012). Atlantic salmon type I IFN subtypes show differences in antiviral activity and cell-dependent expression: Evidence for high IFNb/IFNc-producing cells in fish lymphoid tissues. J Immunol 189, 59125923.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., and Kumar, S. (2011). MEGA5: Molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28, 27312739.

Thomas, C., Moraga, I., Levin, D., Krutzik, P.O., Podoplelova, Y., Trejo, A., Lee, C., Yarden, G., Vleck, S.E., Glenn, J.S., et al. (2011). Structural linkage between ligand discrimination and receptor activation by type I interferons. Cell 146, 621632.

Thompson, J.D., Gibson, T.J., and Higgins, D.G. (2002). Multiple sequence alignment using ClustalW and ClustalX. In Current Protocols in Bioinformatics, John Wiley & Sons, Inc.

Vassileva, G., Chen, S.C., Zeng, M., Abbondanzo, S., Jensen, K., Gorman, D., Baroudy, B.M., Jiang, Y., Murgolo, N., and Lira, S.A. (2003). Expression of a novel murine type I IFN in the pancreatic islets

Protein Cell

&

Higher Education Press and Springer-Verlag Berlin Heidelberg 2013 May 2013 | Volume 4 | Issue 5 | 391

RESEARCH ARTICLE

Lei Xu et al.

induces diabetes in mice. J Immunol 170, 57485755.

Whaley, A.E., Meka, C.S., Harbison, L.A., Hunt, J.S., and Imakawa , K.

(1994). Identication and cellular localization of unique interferon mRNA from human placenta. J Biol Chem 269, 1086410868. Woelk, C.H., Frost, S.D.W., Richman, D.D., Higley, P.E., and Kosak o-

vsky Pond, S.L. (2007). Evolution of the interferon alpha gene family in eutherian mammals. Gene 397, 3850.

Xi, Y., Day, S.L., Jackson, R.J., and Ranasinghe, C. (2012). Role o f novel type I interferon epsilon in viral infection and mucosal immunity. Mucosal Immunol 5, 610622.

Protein Cell

&

392 | May 2013 | Volume 4 | Issue 5 Higher Education Press and Springer-Verlag Berlin Heidelberg 2013