Identification and characterization of two CD4 alleles in Microminipigs

Background We previously identified two phenotypes of CD4+ cells with and without reactions to anti-pig CD4 monoclonal antibodies by flow cytometry in a herd of Microminipigs. In this study, we analyzed the coding sequences of CD4 and certified the expression of CD4 molecules in order to identify the genetic sequence variants responsible for the positive and negative PBMCs reactivity to anti-pig CD4 monoclonal antibodies. Results We identified two CD4 alleles, CD4.A and CD4.B, corresponding to antibody positive and negative, respectively, by nucleotide sequencing of PCR products using CD4 specific primer pairs. In comparison with the swine CD4 amino-acid sequence [GenBank: NP_001001908], CD4.A had seven amino-acid substitutions and CD4.B had 15 amino-acid substitutions. The amino-acid sequences within domain 1 of CD4.B were identical to the swine CD4.2 [GenBank: CAA46584] sequence that had been reported previously to be a modified CD4 molecule that had lost reactivity with an anti-pig CD4 antibody in NIH miniature pigs. Homozygous and heterozygous CD4.A and CD4.B alleles in the Microminipigs herd were characterised by using the RFLP technique with the restriction endonuclease, BseRI. The anti-pig CD4 antibody recognized pig PBMCs with CD4.AA and CD4.AB, but did not recognized those with CD4.BB. We transfected HeLa cells with the FLAG-tagged CD4.A or CD4.B vectors, and certified that transfected HeLa cells expressed FLAG in both vectors. The failure of cells to react with anti-CD4 antibodies in CD4.B pigs was associated to ten amino-acid substitutions in domain 1 and/or one amino-acid substitution in joining region 3 of CD4.B. We also found exon 8 was defective in some CD4.A and CD4.B resulting in the loss of the transmembrane domain, which implies that these CD4 proteins are secreted from helper T cells into the circulation. Conclusions We identified that amino-acids substitutions of domain 1 in CD4.B gave rise to the failure of some CD4 expressing cells to react with particular anti-pig CD4 monoclonal antibodies. In addition, we developed a PCR-RFLP method that enabled us to simply identify the CD4 sequence variant and the positive and negative PBMCs reactivity to our anti-pig CD4 monoclonal antibodies without the need to use flow cytometric analysis. Electronic supplementary material The online version of this article (doi:10.1186/s12917-016-0856-8) contains supplementary material, which is available to authorized users.


Background
The CD4 molecule is a cell-surface glycoprotein receptor expressed by helper T cells, monocytes, macrophages, and dendritic cells; and its structure consists of four immunoglobulin-like domains (D 1 to D 4) as part of the extracellular domain, a transmembrane domain, and a cytoplasmic tail [1,2]. The extracellular domain binds to the monomorphic region of MHC class II to increase the affinity of the T cell receptor to the antigen peptide-MHC class II complex [3,4]. The cytoplasmic portion of CD4+ recruits tyrosine kinase, Lck, and the kinase enhances signal transduction in T cell activation [5,6].
Microminipigs are extra-small and novel miniature pigs developed for biomedical research in Japan [7]. Recently, swine leukocyte antigen (SLA) haplotypes were assigned in a herd of Microminipigs in order to further investigate their immunological characteristics [8] during disease and infections. In the process of analyzing helper T cell function, we found that some pigs had CD4+ cells that could not be detected by flow cytometry while using three anti-pig CD4 monoclonal antibodies of the clones 74-12-4, MIL17, and PT90A [9]. The pedigree analysis indicated that the CD4-undetectable trait might be recessive, suggesting gene variation [9]. Failure of CD4 cells to react with an anti-pig CD4 antibody was reported previously in the NIH miniature swine [10] and the presence of partial nucleotide sequences and 10 amino-acid substitutions in exon 3 and 4 of two kinds of CD4 alleles (CD4.1, CD4.2) in these miniature swine might be the cause of helper T cells not reacting with the anti-pig CD4 antibody [11]. On the other hand, because Microminipigs have no consanguinity with NIH miniature swine, the cause for the failure of CD4 cells to react with the anti-pig CD4 antibody in Microminipigs might be different from NIH miniature swine. Hence, we need to clarify the variations in the CD4 nucleotide and amino acid sequences for the positive and negative antibody phenotypes in Microminipigs.
In this study, in order to clarify the reasons for the failure of the anti-pig CD4 antibodies to react with and detect peripheral CD4+ cells and to assess whether sequence variations within the CD4 molecules of Microminipigs might cause immunological alterations, we 1) sequenced and analyzed the coding sequence (CDS) of CD4 using genomic DNA and reverse transcribed (RT)-PCR products of CD4 mRNA in Microminipigs,, 2) developed a simple PCR-RFLP method to identify the CD4 sequence variant and the positive and negative PBMCs reactivity to anti-pig CD4 monoclonal antibodies, and 3) examined the expression of the CD4 alleles transfected into HeLa cells.

Microminipigs
Microminipigs were raised in a conventional environment at Fuji Micra Inc. (Fujinomiya, Japan) or Gifu University. This study was carried out along the Gifu University Laboratory Animal Guidelines.

Flow cytometric analysis
Flow cytometry was performed as previously [9]. Briefly, peripheral blood mononuclear cells (PBMCs) of 231 Microminpigs were isolated using Lymphoprep (Axis Shield, Oslo, Norway), stained with a FITC-conjugated anti-pig CD4a antibody (clone 74-12-4, BD Biosciences, San Jose, CA), and analyzed using FACSCalibur (BD Biosciences) to classify the pigs with and without CD4 affinity for the 74-12-4 antibody. The data was analyzed with FlowJo version 7.6.5 software (FlowJo, Ashland, OR). The antibody reactivity with the CD4 protein was measured as the MFI (median of fluorescence intensity) of CD4+ cells in PBMCs.

Direct sequencing of CD4 coding region
Genomic DNA was extracted from peripheral blood, pieces of tail or ear tissues of 11 Microminipigs (reactive CD4: six pigs, non-reactive CD4: five pigs) using a Wizard Genomic DNA Purification Kit (Promega Corporation, Madison, WI). CD4 gene-specific primer pairs for amplification and sequencing of the coding region (exons 2 to 10) were designed as shown in Table 1   CD4 genotyping by PCR-restriction fragment length polymorphism (PCR-RFLP) method The PCR-RFLP technique in association with the restriction enzyme BseRI was used to identify and differentiate between the two CD4 alleles. PCR amplification was performed on genomic DNA to amplify CD4 exon 3, and the PCR products were digested with an enzyme, BseRI (New England Biolabs Inc., Ipswich, MA). The allele-specific bands were analyzed by 2 % agarose gel electrophoresis. We determined the hereditary pattern of CD4 alleles of sibs from heterozygous parents by using the PCR-RFLP method to genotype 64 piglets, 35 males and 29 females, born from 17 matings. In addition, CD4 genotypes and phenotypes were assigned in 143 Microminipigs by the PCR-RFLP methods and flow cytometry as described above. The percentage and MFI of CD4+ cells in PBMCs were also compared between the two CD4 genotypes.

Detection of CD4 mRNA by analyzing RT-PCR products
Peripheral blood samples of Microminipigs were collected into Paxgene Blood RNA tubes (PreAnalytiX, Hombrechtikon, Switzerland). Total RNA was extracted using a PAXgene Blood RNA Kit (PreAnalytiX). Complementary DNA (cDNA) was synthesized from oligo dT primers using total RNA and the reverse transcriptase kit, ReverTra Ace (TOYOBO, Osaka, Japan).
To characterise the expressed CD4 mRNA, RT-PCR amplification was performed between exon 1 and 4 or exon 1 and 5 by using two specific primer pairs ( Table 2) that were designed from the nucleotide sequence information of two CD4 alleles, CD4.A and CD4.B, obtained from Microminpigs and the swine CD4 reference sequence [GenBank: NM_001001908]. RT-PCR was performed using the same conditions as those used for sequencing amplification. The RT-PCR products were digested with BseRI, and electrophoresed in 2 % agarose gel.

Transfection of two kinds of CD4 alleles into HeLa cells
We chose HeLa cells for the analysis of CD4 expression analysis in Microminipigs because the same cells have been used for the analysis in human [12,13]. To verify the differences in the antibody reactivity of CD4 alleles, HeLa cells were transfected with the FLAG-tagged CD4 vectors (CD4-FLAG) constructed by adding FLAG to C-terminus of the two different CD4 alleles. PBMCs were isolated from two pigs genotyped to two types of CD4 homozygotes using the Lymphoprep kit (Axis Shield) and their total RNAs were extracted with ISOGEN (NIP-PON GENE, Tokyo, Japan). cDNA was synthesized by SuperScript III First-Strand Synthesis System (Life Technologies) using a sequence specific reverse primer (5′-TCAGGTGAGGGAATAGTTCTTCTGTTGCCG-3′). R T-PCR was performed with PrimeSTAR Max DNA Polymerase (TaKaRa Bio Inc., Otsu, Japan) and a CD4 specific primer pair containing XhoI and NotI recognition sequences and FLAG sequences (Forward: 5′-TATCTC-GAGATGGACCCAGGAACCTCTCT-3′; Reverse: 5′-T ATGCGGCCGCTCACTTGTCATCGTCCTTGTAATC GGTGAGGGAATAGTTCTTCTTCTGTTGC-3′). The RT-PCR products of CD4-FLAG were cloned into the mammalian expression vector pCAGGS [14], and the integrity of the constructed vectors were confirmed by DNA sequencing using the following five primers; Forward 1: 5′-GCAGGGACTTCCTTTGTCCCAAAT-3′; Forward 2: 5′-TATCTCGAGATGGACCCAGGAACCT CTCT-3′; Forward 3: 5′-AGTCACCCTACAGTGCA ATGGAAAG-3′; Reverse 1: 5′-TATGCGGCCGCTCA CTTGTCATCGTCCTTGTAATCGGTGAGGGAATAG TTCTTCTTCTGTTGC-3′; Reverse 2: 5′-TGTCCTTC CGAGTGAGAGACACAA-3′. The constructed plasmids were transfected into HeLa cells using Lipofectamine 2000 (Life Technologies) according to the manufacturer's instructions. After culturing for 20 h, the transfected HeLa cells were stained with a rabbit anti-FLAG polyclonal antibody (Sigma-Aldrich, St. Louis, MO) followed by the Alexa Fluor 488 goat anti-rabbit IgG antibody (Life Technologies). The PE-conjugated mouse anti-pig CD4a monoclonal antibody (clone 74-12-4, SouthernBiotech, Birmingham, UK) was used for the CD4 molecule detection. The cells were also stained with Hoechst 33342 (ImmunoChemistry Technologies, Bloomington, IN). Fluorescence imaging was conducted using scanning laser confocal microscope FV1000-D IX81 (Olympus, Tokyo, Japan).

Statistical analysis
In the hereditary analysis, the observed and theoretical values were assessed by the chi-squared test using Excel 2007 (Microsoft, Seattle, WA) with an add-in software Statcel 3 (OMS, Tokorozawa, Japan). Theoretical values were determined on the basis of the Punnett square. A difference of P < 0.05 was considered as significant.

Results
PCR amplification of CD4 gene sequences between exons 2 and 10 was performed on 11 Microminipigs, six pigs that were CD4 antibody reactive and five pigs that were CD4 antibody unreactive. DNA sequencing and analysis of the 11 PCR products identified three  [11].
In comparing the derived CD4 protein sequences with the swine CD4 amino-acid reference sequence [Gen-Bank: NP_001001908], the CD4.A and CD4.B protein sequences had seven and 15 amino-acid substitutions, respectively, in the regions of exons 2 to 10 ( Fig. 1, Table 4). In CD4.A, there was one amino-acid substitution in three of the four extracellular domains as well as in the joining regions 1 and 4, and two amino-acid substitutions in the transmembrane domain. In CD4.B, there were ten amino-acid substitutions in domain 1, one in domain 3, one each in joining regions 3 and 4, and two in the transmembrane domain, some of which may change the polarity or charge of the amino-acid side chains. There was no amino-acid substitution in the cytoplasmic region of either CD4.A or CD4.B.
Three CD4 genotypes in Microminipig herd were assigned as CD4.AA, AB, and BB by the PCR-RFLP method using BseRI (Fig. 2). The restriction enzyme patterns of CD4.AA, AB and BB showed a single band (366 bp), three bands (366, 260, and 106 bp), and two bands (260 and 106 bp), respectively. The matings of 17 pairs of heterozygous parents revealed that the inheritance pattern of CD4 genotypes was autosomal ( Table 5). As shown with the flow cytometry results in Table 6, PBMCs with CD4.AA and AB reacted with the antibody clone 74-12-4. In contrast, PBMCs with CD4.BB were unreactive with the antibody. The MFI of CD4.AB was approximately half the intensity of CD4.AA, even though the percentage of CD4+ cells in PBMCs were not different between the two CD4 alleles (Fig. 3).
CD4 gene expression was analyzed by RFLP using the RT-PCR sequence products that were amplified between exons 1 and 4 (Fig. 4a) or between exons 1 and 5 (Fig. 4b). RFLP distinguished the genotypes CD4.AA, CD4.BB and CD4.AB in both cases. In Fig. 4a, the RT-PCR products were detected as a single 400 bp-band by electrophoresis. After BseRI digestion, the single band remained in CD4AA, whereas two digested bands of 303 bp and 97 bp were obtained in CD4.BB. Combinatorial band patterns of CD4.AA and CD4.BB were observed in CD4.A/B. In Fig. 4b, the undigested RT-PCR products were detected as a single 595 bp band. After the BseRI digestion, the single 595 bp band remained in CD4.AA, whereas two digested bands of 300 bp and 295 bp were obtained in CD4.BB, and the combinatorial band patterns of CD4.AA and CD4BB were observed in CD4.A/B. Consequently, these results suggested that PBMCs with heterozygous CD4 genotype coexpressed CD4.A and CD4.B alleles at the mRNA level.
In validating the expression vector sequences, the insertion sequences of CD4.A-FLAG and CD4.B-FLAG were found to be identical to the genomic exon sequences described above (Additional file 1) except for the added FLAG sequence. Moreover, we also found a spliced form that lacked the CD4 exon 8 in both of the two CD4 alleles. These spliced forms with the exon 8 deficiency gave rise to a stop codon at the N-terminus of transmembrane domain as a result of a frameshift from the beginning of the exon 8 region, whereas amino-acid sequences of the external domains in the spliced forms were identical to those of the CD4.A and CD4.B derived from the nucleotide sequencing using genomic DNA (Fig. 5). Therefore, we used the constructs with complete sequences of CD4-FLAG for expression in HeLa cells. These alternative spliced forms were submitted to DDBJ (http://www.ddbj.nig.ac.jp) as  Figure 6 shows the transient expression of CD4-FLAG without the exon 8 deficiency in HeLa cells. The CD4.A and FLAG proteins in CD4.A-FLAG were detected with the anti-pig CD4 antibody and anti-FLAG antibody. In contrast, the CD4.B in CD4.B-FLAG was unreactive with the anti-pig CD4 antibody even though FLAG was detected with the anti-FLAG antibody. These results show that we expressed the CD4.B protein in HeLa cells, but that we could not detect it with the anti-CD4 antibody.

Discussion
The CD4 exonic sequences of the Microminipig CD4 gene were analyzed using the DNA isolated from CD4+ cells that were CD4-reactive and non-reactive to the 74-12-4 antibody, and two corresponding alleles, CD4.A and CD4.B, respectively, were identified. Although the CD4 gene is thought to be highly conserved, CD4 polymorphisms were reported previously in human, bovine, and ovine [15][16][17]. Also, two CD4 partial allelic sequences were reported in NIH miniature swine (CD4.1: X65629;  However, the Microminipig breed is the result of crosses with western breeds, but it has no consanguinity with the NIH miniature swine [7]. Although the CD4.B of Microminipigs and CD4.2 of NIH miniature swine might have a co-ancestor, the origins of CD4.B and CD4.2 remain uncertain. In comparing the amino-acid sequence alignments of CD4.A and CD4.B with the swine CD4 amino-acid reference sequence [GenBank: NP_001001908], we identified one amino-acid substitution in domain 4 of the extracellular region of CD4.A that involve alterations in the charge of the amino-acid side chain, and consequently a structural change in this domain of CD4. Domains 3 and 4 of CD4 play an important role in interacting with the T cell receptor-CD3 complex to influence signal transduction in T cell activation and function [18]. Therefore, the amino-acid substitution on domain 4 of CD4.A might affect signal transduction and T cell function during the interaction of CD4 with the T cell receptor-CD3  Table 1). The PCR product was digested with BseRI. PCR fragments with genotype of CD4.AA, CD4.AB, and CD4.BB showed single fragment (366 bp), three fragments (366, 260, and 106 bp), and two fragments (260 and 106 bp), respectively The hereditary pattern corresponds to the theoretical value on the basis of autosomal heredity in heterozygous by analysis of chi-squared test (P > 0.05). Hence, the hereditary pattern of CD4 genotypes is autosomal in mode complex. The five amino-acid substitutions observed in domain 1 of CD4.B also alter the polarity or charge of amino-acid side chains and might elicit a structural alteration in this domain of the CD4. The domain 1 of the CD4 combines with the monomorphic region of the MHC molecule in the presentation of antigenic determinants to activate selected lymphocytes [3,4]. Moreover, the amino-acid substitutions in the region of domain 1 in CD4.B correspond to the CDR2-like region of CD4.2 in NIH miniature swine and the CDR2-like region of CD4 in humans that bind to MHC class II molecules and the HIV envelope glycoprotein gp120 [11]. No clinical abnormalities have been observed as yet with the CD4.B genotype in Microminipigs, although the affinity of CD4.B to MHC class II might be different from that of CD4.A [9]. In this regard, the PCR-RFLP technique using BseRI has allowed us to identify the three CD4 genotypes CD4.AA, AB and BB that correlated with CD4 reactivity to anti-pig CD4 antibodies. This simple CD4 genotyping method might be useful for selectively breeding CD4.A or CD4.B homozygous pigs and for developing association studies of immunity to infections and immunologically-related diseases.
Our PCR-RFLP study has demonstrated that both CD4.A and CD4.B are co-expressed in PBMC of heterozygous pigs. Moreover, the study on the expression of CD4-FLAG in HeLa cells confirmed that the amino-acid substitutions in CD4.B were associated with the loss of affinity to anti-CD4 antibody. On the other hand, the seven amino-acid substitutions observed within CD4.A reacted with the anti-CD4 antibody and one or other of them are likely to be linear or sequential epitopes recognized by the anti-CD4 antibody. Therefore, one or more of the ten amino-acid substitutions in domain 1 and/or the one amino-acid substitution in joining region 3 of CD4.B, which are not found in CD4.A, may have replaced the antigenic determinant and caused the lose of affinity for the anti-pig CD4 antibodies.
The detection of reduced levels of MFI in CD4 cells with CD4.AB in the Microminipigs appears to be due to the anti-CD4 antibody reacting with CD4.A, but not with the CD4.B molecules, if both types of CD4 proteins are coexpressed on the surface of helper T cells. Moreover, in CD4-FLAG insertion sequencing, the CD4.A and CD4.B alleles also had exon 8 deficiency forms that lacked the subsequent of transmembrane domain in the deduced amino-acid sequences. Thus, if the CD4 transcripts are without a transmembrane domain then the translated proteins might be secreted into the serum rather than be bound within the cellular membrane. The alternative spliced forms of CD4 CDS in swine have not been reported previously even though there are such variants registered in GenBank [GenBank: XM_005652591, GenBank: XM_005652592, GenBank: KC333254, GenBank: AY515293]. However, the lack of CD4 transmembrane region was reported for a mutant mouse model that secreted soluble CD4 without expression of membrane-bound CD4 [19]. This mouse model was used to show that soluble CD4 impaired a delayed-type hypersensitivity response by inhibiting IFN-γ production, and prohibited over-activation of CD4+ T cells by competitive inhibition of the binding of CD4 on the T-cell surface to MHC class II [20]. So, if the exon 8 deficiency forms are translated to the protein, the secreted CD4 might be also associated with prohibiting over-activation of CD4+ T cells in swine. Thus, further studies are needed to elucidate the significance of the expression of CD4 exon 8 deletions.
In NIH miniature swine, the functional differences of CD4 between CD4.1 and CD4.2 were investigated, but no differences were detected in antibody production against staphylococcal nuclease immunization and in the allogeneic mixed lymphocyte reaction [21]. In this regard, additional studies will be needed in Microminipigs to elucidate the functional significance or immunological importance of polymorphisms of the CD4 gene including the possible alternative spliced forms of the expressed gene. Because the CD4 molecule interacts with the MHC class II complex in antigen recognition [3,4] the polymorphisms of both CD4 and MHC class II will need to be considered in future studies. The Microminipigs with defined SLA haplotypes [8] could be a useful animal model for further research on the interaction between CD4 allomorphs and MHC molecules in disease, infection and transplantation studies.  Table 2 and the amplified products were digested with BseRI. a After digestion with BseRI, CD4.AA, CD4.AB, and CD4.BB showed a 400 bp-fragment (400 bp), three fragments of 400, 303 and 97 bp, and two fragments of 303 and 97 bp, respectively. b After digestion with BseRI, CD4.AA, CD4.AB, and CD4.BB showed a 595 bp-fragment, three fragments of 595, 300 and 295 bp, and two fragments of 300 and 295 bp, respectively  Expression of the CD4-FLAG vectors in HeLa cells. We expressed only the constructs CD4-FLAG with complete sequences in HeLa cells as described in Fig. 5. a Cells stained with rabbit anti-FLAG antibody followed by Alexa Fluor 488 goat anti-rabbit IgG antibody. b Cells stained with PE conjugated mouse anti-pig CD4 antibody. c Cells stained with the overlay of three fluorophore signals; anti-FLAG antibody, anti-pig CD4 antibody, and Hoechst 33342