In a recent report, H-T Lieu et al.1 created a Reg2/RegIIIβ gene deficiency (marked as Reg2−/−) through homologous recombination. The knockout mice were normal except for impaired hepatocyte survival and regeneration when challenged by induced hepatitis or hepatectomy. Thus, in response to Fas-induced acute hepatitis, Reg2−/− mice had decreased survival; after a two-thirds hepatectomy, they exhibited increased mortality, delayed liver mass restoration, and decreased bromodeoxyuridine incorporation into their DNA. The authors even presented evidence that endogenous Reg2 gene expression and protein production were induced during liver regeneration. It was concluded that Reg2/HIP/PAP is a critical mitogenic and antiapoptotic factor for the liver. Although we welcome the data and interpretation, it is totally unclear on which gene, Reg2 or Reg3β, was actually targeted that affected liver regeneration.
According to the article, “Reg2, also called RegIIIβ, is the mouse homolog of human hepatocarcinoma-intestine-pancreas/pancreatic-associated protein (HIP/PAP), whose cDNA has been cloned in different species under different names: human HIP/PAP or Reg3A; rat Reg2, PAPI, or peptide 23; mouse Reg2/RegIIIβ; and hamster islet neogenesis associated protein or INGAP, respectively”, and the target gene to these underlined proteins could be Reg2, Reg3α, Reg3β, or INGAP,1 respectively. This is impossible because they represent independent genes encoding unique proteins, assuming only one gene was targeted. Similar confusing terminology has been used in another article from the same laboratory where HIP/PAP was referred to as human Reg-2, in the abstract, but Reg2 has only been found in mice so far.2
We have reason to believe that the actual target was mouse Reg3β (also known as Reg IIIβ; or PAP, PAP1 and HIP in other species).3–5 As shown in Fig. 1, the polymerase chain reaction (PCR) primers used to genotype the knockout offspring were perfectly matched to the genomic sequence of mouse Reg3β gene (access number D63360), which would generate a product of 687 base pairs, similar to the 600 base pairs reported.1 The primer sequences were also found in the mouse Reg3β cDNA (NM_011036) at nucleotide positions 295-315 and 376-395, respectively. Moreover, the same genomic DNA (D63360) has a unique EcoR V site at position 2175-2180 of the intron 2, exactly as illustrated in figure 1A of the article.1 The antibody used to detect Reg2 using immunohistochemistry was raised against human HIP/PAP which should correspond to mouse Reg3β.1, 2
We also have evidence to exclude other Reg genes as the target. The mouse genome has been sequenced and much information is available from Web sites such as www.ncbi.nlm.nih.gov or www.informatics.jax.org. In addition to Reg3β, we explored 3 other possible candidates, as listed in Table 1. The PCR primers were not matched to any other genomic or cDNA sequences (by access numbers in Table 2). The Reg2 locus has no EcoR V site, Reg2 ispredominately expressed in the pancreatic islets,6 and a homologous Reg2 gene has not been found in human or rats. The Reg3α locus has an EcoR V site in intron 4, instead of intron 2 as reported;1 the Reg3δ gene (mouse homolog of INGAP, also known as INGAP-rp) has an EcoR V site in intron 1, instead of intron 2.
|Gene (Reference)||Access No.||Size in kb||Number of EcoR V||Position||Location|
|Reg3α (5)||D63358||4.3||1||3352||Intron 4|
|Reg3β (5)||D63360||4.3||1||2175||Intron 2|
|Reg3δ (8)||AB035205||4.5||1||1805||Intron 1|
|Mouse Gene (Reference)||Orthology||Amino Acids||GenBank Access Numbers||Chromosome Location|
|Reg1 (7, 11,12)||Reg, PTP, PSP||165||D14010, NM_009042, P43137||6, Cytoband C-D|
|Reg2 (7)||PTP2, PSP2, lithostathine2||173||D14011, NM_009043, Q08731||6, Cytoband C|
|Reg3α (5)||PAP2||175||D63358, NM_011259, O09037||6, Cytoband C|
|Reg3β (3–5)||PAP, PAP1, HIP||175||D63360, NM_011036, P35230||6, Cytoband C|
|Reg3γ (5)||PAP3||174||D63362, NM_011260, O09049||6, Cytoband C|
|Reg3δ (8,13)||INGAP-rp, INGAP||175||AB035205, NM_013893, Q9QUS9||6, Syntenic|
|Reg4 (14–16)||RELP||157||67709, NM_026328, Q9D8G5||3, Cytoband F3|
The current terminology used for mammalian Reg family genes/proteins is very confusing and should be unified through debates. According to our search, there seem to be 7 Reg family genes in the mouse (Table 2). Similar and more detailed classifications need to be done for human and rat in the future. Thus, unless the authors provide further evidence, we urge appropriate corrections to be made to this and the previous articles, where Reg2 and HIP/PAP should be referred to as Reg3β.