tein raises very interesting questions about the structure of the virus. One possibility is that the virus might usurp a cellular protein (or a protein encoded by a different virus) in In this issue of Hepatology, several articles address the lieu of encoding its own capsid protein (much like hepatitis clinical consequences of infection with a novel human flavidelta virus uses the envelope protein of hepatitis B virus virus, variously called ''hepatitis G virus,'' ''GB virus C'' [HBV]). Alternatively, GBV-C could form a capsid-less parti-(GBV-C), or ''hepatitis virus GB-C.'' The discovery of this cle. While not impossible, this would be without precedent virus, reported independently by Simons et al. in 1995 1 and among naturally occurring RNA viruses. These aspects of Linnen et al. in 1996, 2 raised expectations that an elusive, GBV-C point strongly to the conclusion that GBV-C is not sixth human hepatitis virus had finally been identified. Inisimply a variant of HCV, but a virus that differs profoundly tial enthusiasm was fueled by the inclusion of ''hepatitis'' in from HCV in ways that are likely to affect both its pathogenicthe names given this agent by some investigators. However, ity and epidemiology. few data have since been presented that suggest that this GBV-C was discovered subsequent to the identification of virus actually causes either acute or chronic liver disease in GB viruses A (GBV-A) and B (GBV-B) in the blood of tamahumans.
rins that had been inoculated with material passaged from As most readers know, GBV-C is a positive-stranded RNA a surgeon with hepatitis. 4 GBV-A has a 5 genome structure virus with a genome approximately 9.4 kb in length that and organization resembling that of GBV-C, with which it is contains a single long open reading frame (ORF). 1,2 The ORF relatively closely related at the nucleotide sequence level. It encodes a lengthy polyprotein which includes two putative appears to be a tamarin homolog of the human GBV-C. An envelope glycoproteins near its amino terminus and helicase, additional virus, closely related to GBV-A, has been recovered protease, and RNA-dependent RNA polymerase motifs furfrom New World owl monkeys. 5 The data suggest that GBVther toward its carboxy terminus. These features resemble A and GBV-C may have evolved from an ancient virus that the genome organization of hepatitis C virus (HCV) and have infected an ancestor common to both New World and Old led to the informal classification of GBV-C within the family World primates before the separation of the continents. Flaviviridae. Although the protease/helicase domain of GBV-GBV-B remains an enigma, as additional isolates of this C shares clear homology with the HCV NS3 protein, there
virus have yet to be reported. However, it is interesting to are two unique attributes of GBV-C that dramatically distinnote that GBV-B encodes a nucleocapsid protein and has a guish it from other flaviviruses. First, the 5 nontranslated 5NTR secondary structure with many features similar to RNA (5 NTR) of GBV-C is relatively lengthy (ΓΊ540 nts) and that of HCV. Early studies showed that GBV-B is a hepatoshares neither significant primary nor secondary RNA structropic agent, capable of inducing hepatitis in tamarins. 5 Simiture with the 5NTR of HCV. 3 Like HCV, the 5NTR of GBVlar studies failed to confirm the liver as a site of replication C contains an internal ribosome entry site (IRES) that is for GBV-A, however, and the tissue specific tropisms of GBVcapable of directing cap-independent translation of the poly-A and GBV-C remain undefined. The amplification of GBVprotein. 3 However, this IRES has extraordinarily weak activ-C sequences from liver tissue by polymerase chain reaction, ity compared with the HCV IRES. Furthermore, structural as reported by Lo Β΄pez-Alcorocho et al. 6 in this issue of Hepatolcomparisons suggest that the 5NTRs of GBV-C and HCV ogy, does not resolve this question, because it is difficult to may not be phylogenetically related, raising the interesting imagine how even extensive washing of a biopsy sample could possibility that an ancient RNA recombination event may effectively remove blood permeating the hepatic sinusoids. have set these viruses on their separate evolutionary path-What, then, is the evidence that GBV-C causes liver disways.
ease in humans? Linnen et al. 2 claimed an ''association'' be-More significant, perhaps, is the fact that the GBV-C getween this virus (which they named ''hepatitis G virus'') and nome does not appear to encode a nucleocapsid protein analoboth acute and chronic hepatitis in humans. However, the gous to the capsid protein of HCV. Carefully performed analyword ''associated'' has several possible interpretations. It is ses of the protein products obtained following translation of usually taken to denote a statistically significant connection GBV-C RNA in rabbit reticulocyte lysates indicate that transbetween two entities, but not necessarily an etiologic conneclation commences at an AUG codon located immediately uption. An association may also be epidemiological, if two clinistream of the signal sequence for the first envelope glycoprocal entities share common risks for their acquisition (for extein. 3 While some strains of GBV-C do have patent ORFs ample, acquired immune deficiency syndrome and HBV that extend 5 of this AUG, there are as of yet no data that infection). The proponderance of available data suggest just suggest that this potential coding sequence is expressed.
such an association between GBV-C infection and hepatitis Moreover, the 5NTR secondary structure is conserved in in humans. these strains, suggesting that all GBV-C strains are likely to Linnen et al. 2 presented two lines of evidence for an associinitiate translation under direction of functionally similar ation between GBV-C infection and acute liver disease in humans. First, GBV-C RNA was identified in sera from 2 of 12 patients with acute, posttransfusion non-ABC hepatitis.
Abbreviations: GBV-C, GB virus C; ORF, open reading frame; HCV, hepatitis C virus;
It seems likely that this rate of GBV-C infection simply re-5NTR, 5 nontranslated RNA; IRES, internal ribosome entry site; HBV, hepatitis B virus;
flects the large number of units received by these transfused GBV-A, GB virus A; GBV-B, GB virus B; ALT, alanine transaminase; HOFV, human orphan individuals (9 and 14 units in the two infected persons), and flavivirus.
the high rate of GBV-C infection that these authors docu-