Geographic distribution and genetic diversity of Whitewater Arroyo virus in the southwestern United States.

The purpose of this study was to extend our knowledge of the geographic distribution and genetic diversity of the arenavirus(es) associated with Neotoma species (woodrats) in the southwestern United States. Infectious arenavirus was recovered from 14 (3.3%) of 425 woodrats. The virus-positive species included N. albigula in New Mexico and Oklahoma, N. cinerea in Utah, N. mexicana in New Mexico and Utah, and N. micropus in Texas. Analyses of viral nucleocapsid protein gene sequence data indicated that all the isolates were strains of the Whitewater Arroyo virus, an arenavirus previously known only from northwestern New Mexico. Analyses of the sequence data also indicated that there can be substantial genetic diversity among strains of Whitewater Arroyo virus from conspecific woodrats collected from different localities and substantial genetic diversity among strains from different woodrat species collected from the same locality.

The purpose of this study was to extend our knowledge of the geographic distribution and genetic diversity of the arenavirus(es) associated with Neotoma species (woodrats) in the southwestern United States. Infectious arenavirus was recovered from 14 (3.3%) of 425 woodrats. The virus-positive species included N. albigula in New Mexico and Oklahoma, N. cinerea in Utah, N. mexicana in New Mexico and Utah, and N. micropus in Texas. Analyses of viral nucleocapsid protein gene sequence data indicated that all the isolates were strains of the Whitewater Arroyo virus, an arenavirus previously known only from northwestern New Mexico. Analyses of the sequence data also indicated that there can be substantial genetic diversity among strains of Whitewater Arroyo virus from conspecific woodrats collected from different localities and substantial genetic diversity among strains from different woodrat species collected from the same locality.

Research
Corning, NY). The inoculum was incubated on the cell monolayer at 37°C for 60 minutes; then the monolayer was overlaid with 7.0 mL of a minimum essential medium containing Earle's salts, 1.5 mg/mL sodium bicarbonate, 2% v/v heat-inactivated (56ºC for 30 minutes) fetal bovine serum, 0.29 mg/mL L-glutamine, 100 U/mL penicillin G, 100 µg/mL streptomycin sulfate, and 100 U/mL nystatin. The cell culture was maintained at 37°C in a humidified atmosphere of 5% CO 2 in air for 13 days. Half the culture medium was replaced with fresh maintenance medium on day 6 or 7 after inoculation. Cells were scraped from the monolayer on day 13 after inoculation and coated onto 12-well glass microscope slides (Cel-Line Associates, Inc., Newfield, NJ). The cell spots were air-dried, fixed in cold acetone, and then tested for arenaviral antigen by using an indirect fluorescent antibody test (12). In that test, cell spots were stained with a hyperimmune mouse ascitic fluid prepared against the WWA virus prototype strain AV 9310135, and mouse immunoglobulin G (IgG) bound to cell-associated arenaviral antigen was detected by using a goat anti-mouse IgG fluorescein isothiocyanate conjugate (Cappel Laboratories, West Chester, PA).

Genetic Characterization of Viral Isolates
The nucleotide sequence of a fragment of the N protein gene of each of 12 isolates was determined. Four of the 12 isolates were from the spleens and kidneys of two animals, rodents 62425 and 62439 (Table 1). Total RNA was extracted from monolayers of infected Vero E6 cells by using TRIzol Reagent (Life Technologies, Inc., Grand Island, NY). Reverse transcription of RNA from isolates AV 96010149, AV 96010151, AV 96010025, and AV 96010024 was carried out by using Superscript II RTase (Life Technologies) in conjunction with oligonucleotide ARE-3'END (14). This oligonucleotide apparently is complementary to the 19-nt fragment at the 3' terminus of the S genomic segment of all arenaviruses. Polymerase chain reaction (PCR) amplification of the firststrand cDNA was carried out by using Taq DNA polymerase (Promega Corp., Madison, WI) in conjunction with oligonucleotides 1010C and NW1696R (2-3), which flank a 616-nt region of the N protein gene of WWA virus prototype strain AV 9310135 (12). Reverse transcription and PCR (RT-PCR) amplification of a fragment of the N protein gene of each of the eight other isolates was carried out by using the Access RT-PCR Kit (Promega Corp.) in conjunction with oligonucle-otides AVNP1 (5'-CCCTTCTTYTTNYTCTTRATGACTA-3') and AVNP2 (5'-GGKAGRGCNTGGGAYAACAC-3'). AVNP1 and AVNP2 flank a 518-nt region in the fragment of the WWA virus N protein gene that is amplified by using oligonucleotides 1010C and NW1696R. They were designed based on N protein gene sequence data for the WWA virus prototype strain AV 9310135 (GenBank Accession No. U52180), WWA virus strains AV 96010149, AV 96010151, AV 96010025, and AV 96010024, TAM virus strain W-10777 (U43690), and PIC virus strain An 3739 (K02734). Size separation of PCR products was done by agarose gel electrophoresis; the products of the expected size were purified from gel slices by using QIAquick Gel Extraction Kit (Qiagen, Inc., Valencia, CA). One strand of each 1010C-NW1696R PCR product was sequenced directly by using the dye termination cycle sequencing technique (Applied Biosystems, Inc., Foster City, CA) in conjunction with oligonucleotide 1010C. The sequence of the other (i.e., complementary) strand of each of these products was determined by cloning the PCR product in the TA cloning vector PCRII (Invitrogen Corp., Carlsbad, CA) and then using a plasmid-specific oligonucleotide (M13) to initiate the cycle sequencing reaction. Both strands of the AVNP1-AVNP2 PCR products were sequenced directly by using the same oligonucleotides that were used to prime the RT-PCR, i.e., AVNP1 and AVNP2. The 12 nucleotide sequences generated in this study were deposited with the GenBank nucleotide sequence database under Accession Nos. AY012710-AY012721.

Data Analysis
The analyses of nucleotide sequence data were restricted to the 518-nt fragment of the WWA virus N protein gene that is flanked by oligonucleotides AVNP1 and AVNP2. The

Research
CLUSTAL W1.7 (15) was used to construct an alignment of the predicted amino acid sequences, and the computer program TransAlign (16) was used to generate a multiple nucleotide sequence alignment from the amino acid sequence alignment. Pairwise genetic distances were computed by using the p distance model as implemented in the computer program MEGA, version 1.02 (17). Percent sequence identities were calculated by subtracting the genetic distances from 1.0 and multiplying by 100. Phylogenetic analysis was carried out on the multiple amino acid sequence alignment by using the neighbor-joining method (gamma model, alpha = 2) as implemented in MEGA, version 1.02. Bootstrap support (18) for the results of the phylogenetic analysis was based on 500 pseudoreplicate datasets generated from the original multiple amino acid sequence alignment.
The The nucleotide sequences of the isolates from rodent 62425 (one isolate each from kidney and liver; strains AV 96010151 and AV 98360019, respectively) were identical. In contrast, the nucleotide sequences of the isolates from rodent 62439 (again, one isolate each from kidney and liver; strains AV 96010154 and AV 98360020, respectively) were 99.6% identical. Further study is needed to determine whether the differences between the isolates from rodent 62439 represent the coexistence of multiple virus genotypes (alleles) in the same rodent. An alternative explanation is that the sequence differences are the result of adaptation of the isolates to growth in cultured (Vero E6) cells or manipulation of viral nucleic acid extracted from cultured cells.
Nucleotide and amino acid sequence identities among WWA virus prototype strain AV 9310135 and 12 other

Research
isolates from Neotoma rodents ranged from 74.7% to 100.0% and 84.9% to 100.0%, respectively ( Phylogenetic analysis of N protein amino acid sequence data indicated that isolates from Neotoma rodents represent a phylogenetic lineage (viral species) that is in a sister relationship to the lineage represented by TAM virus ( Figure  2). We concluded that all isolates recovered from the Neotoma rodents were strains of WWA virus.

Discussion
Before the present study, WWA virus was known only from N. albigula in northwestern New Mexico (12). The present work provides unequivocal evidence that the virus also is naturally associated with N. cinerea, N. mexicana, and N. micropus, and that it occurs in Utah, central New Mexico, Oklahoma, and Texas. The recovery of WWA virus strains AV 98490013 and TVP-6083 from N. albigula is the first evidence that a Tacaribe complex virus occurs in Oklahoma. Likewise, the recovery of strains AV A0400174 and AV A0400212 from woodrats collected from southern Texas (Chaparral Wildlife Management Area) is the first evidence that N. micropus is naturally associated with a Tacaribe complex virus and that WWA virus occurs in Texas.
In a previous study (13), antibody to an arenavirus was found in N. fuscipes and N. lepida in southern California; N. albigula, N. mexicana, and N. stephensi in Arizona; and N. albigula in southwestern Colorado. Although the results of the present study indicate that WWA virus is geographically widely distributed in association with Neotoma rodents, further work is needed to determine whether the arenavirus associated with Neotoma rodents in California, Arizona, and Colorado is in fact WWA virus.
The results of the present study indicate that there can be substantial genetic heterogeneity among strains of WWA virus from different woodrat species from the same locality and among strains from conspecific woodrats collected from different localities. For example, nucleotide sequence identity between the strains recovered from N. mexicana and N. cinerea from Natural Bridges National Monument (San Juan County, Utah; strains AV 96010024 and AV 96010025, respectively) was 82.8%, and nucleotide sequence identity between strain AV 96010024 and the three strains recovered from N. mexicana collected from the Magdalena Mountains (Socorro County, New Mexico; strains AV 96010149, AV 96010151, and AV 96010154) was from 85.1% to 85.5%. In contrast, nucleotide sequence identity in strains recovered from conspecific rodents collected from the same locality (e.g., strains AV 9310135 and AV 9310040 from N. albigula from Whitewater Arroyo, and strains AV A0400174 and AV Figure 2. Phylogeny of the North American arenaviruses based on a neighbor-joining analysis of nucleocapsid protein amino acid sequence data. Distances and groupings were determined by using the gamma distance algorithm (alpha = 2). Branch lengths are proportional to the gamma distance between amino acid sequences. Numbers indicate the percentage of 500 bootstrap replicates that supported each labeled interior branch. The WWA virus prototype strain AV 9310135 is in bold type. Nmex = Neotoma mexicana, Nalb = N. albigula, Ncin = N. cinerea, Nmic = N. micropus, and Shisp = Sigmodon hispidus.  (3,19,20) suggested that the present-day diversity of the arenaviruses is a product of long-term coevolution of the various viruses with their respective principal rodent hosts. In the present study, WWA viral strains AV 9310135 and AV 9310040 (both from N. albigula, northwestern New Mexico) appeared to be phylogenetically more closely related to strain AV 96010024 (N. mexicana, southeastern Utah) than to strains AV 98490013 and TVP-6038 (both from N. albigula, western Oklahoma). This situation suggests that the present-day association of WWA virus with N. albigula and N. mexicana does not represent a long-term shared evolutionary relationship between virus and rodent species. However, this conclusion assumes that recovery of WWA virus from a rodent represents a principal virus-host relationship. Perhaps some of the virus-positive rodents in the present study were infected by contact with other Neotoma species or even non-Neotoma rodent species.
The geographic range of the genus Neotoma extends from western Canada south to Guatemala, Honduras, and Nicaragua, and includes 33 states in the contiguous United States and 26 of the 32 states in Mexico (21). Thus, if the present-day association of WWA virus with the genus Neotoma represents a long-term shared evolutionary relationship between virus and rodent host, the geographic range of the virus may extend far beyond the southwestern United States. WWA virus recently was associated with several human deaths in California (22). Further study is needed to assess the human health significance of this virus in the southwestern United States and other regions in North America in which woodrats are indigenous.