Visna sequences

Does visna + something = HIV-1?

Sequence comparison can tell us. We find:

HIV is not especially similar to visna, HTLV-I, or BLV. It is not formed by splicing together any of these viruses. HIV is closest to SIV (the monkey virus). Click here for a detailed look at how these viruses are related.

These results totally destroy the theory that HIV was formed by evil scientists recombining some combination of these viruses.

What are these sequences?

For our purposes, you can consider the DNA (or RNA) sequence of a virus to be about 10,000 letters "c", "g", "a", or "t". The sequence can be determined by laboratory techniques. The sequence describes, through the genetic code, the sequence of amino acids that go into the proteins that the virus uses. These protein sequences can be described as several hundred letters, with one letter for each amino acid. The genetic sequence of viruses changes as they mutate, so by comparing the sequences, we can see which viruses are more closely related.

A look at some sequences

NEEAERWVRQNPPGPNVLTVDQIMGVGQTNQQASQANMDQ-------------ARQICLQWVITAL  visna
NEEAAEWDRVHPVHAGPIAPGQMREPRGSDIAGTTSTLQEQIGWM-TNNPPIPVGEIYKRWIILGL  hiv-1
NEEAAQWDIAHPPPAGPLPAGQLRDPRGSDIAGTTSTVQEQLEWIYTANPRVDVGAIYRRWIILGL  siv
****  *    *         *                                  *   * *  *

Note that on this protein fragment HIV and SIV are 70% similar, while visna is only 25% similar. It is clear, though, that all three viruses are related. Note that the sequences all match up at the start and end, but in the middle there is a sequence common to HIV and SIV that is missing from visna. If HIV came from visna, where did this middle sequence come from, and why does it match the SIV sequence closely? On the other hand, if HIV and SIV came from a common ancestor, that would easily explain it.

Let's look at two hypotheses: a) the HIV virus was made by combining bovine leukemia virus (BLV) with sheep visna (VLV) virus

b) HIV, the green monkey virus (SIV), BLV, and VLV all evolved from a common ancestor. HIV and SIV evolved from a close ancestor; BLV and VLV are more distantly related.

First, a bit of background information. The RNA of retroviruses encodes various proteins. Each protein is made up of a chain of particular amino acids, selected from the 20 amino acids, according to the RNA sequence. With sequencing, we can determine the amino acid sequence in these proteins.

Now, based on the two hypotheses, we can make some predictions: If hypothesis a) is true, we would expect the sequence for a protein in HIV to match the sequence of either the corresponding BLV protein or the VLV protein. Alternatively, the sequences could have recombined, so the HIV sequence could consist of parts of the BLV and VLV sequences.

If hypothesis b) is true, we would expect the sequences to be close in some areas and to have evolved apart by mutation in other areas. We would expect the HIV and SIV sequences to be closest, and the HIV sequence to be more different from BLV and VLV.

Now, we can look at the sequences for the POL gene in the four viruses, match up the sequences, and compare them. J. Hein's alignment program finds the HIV and SIV POL proteins are about 60% similar, VLV is about 25% similar to HIV and SIV, and the BLV protein is about 15% similar to the other viruses. These numbers strongly suggest that HIV came from SIV (or a close ancestor), and not BLV or VLV. I don't trust these numbers too much, though, because BLV is sufficiently different that it is hard to tell if the matching is real in a lot of places.

Thus, I've selected a short part of the POL gene where all four viruses match up well, so it is clear the matching is real and not coincidental. (0=HIV, 1=SIV, 2=VLV, 3=BLV)

NTPVFAIKKKDSTKWRKLVDFRELNKRTQDFWEVQLGIPHPAGLKKKKSVT-V--LDVGDAYFSVPLDED  0
NTPVFAIKKKDKTQWRMLVDFRELNKATQDFFEVQLGIPHPAGLQKKKQIT-V--IDIGDAYYSIPLCKE  1
NTPIFCIKKKSG-KWRMLIDFRELNKQTEDLAEAQLGLPHPGGLQRKKHVT-I--LDIGDAYFTIPLYEP  2
NNPVFPVRKPNG-AWRFVHDLRATNALTKPIPALSPGPPDLTAIP--THPPHIICLDLKDAFFQIPVEDR  3
* * *   *     **   * *  *  *        * *                 *  **    *    

FRKYTAFTIPSINNETPGIRYQYNVLPQGWKGSPAIFQSSMTKILEPFRKQNPDIVIYQYMDD  0
FRKYTAFTIPSVNNTGPGIRYQFNCLPQGWKGSPTIFQNTAANILEEIKRHTPGLEIVQYMDD  1
YRQYTCFTMLSPNNLGPCVRYYWKVLPQGWKLSPAVYQFTMQKILRGWIEEHPMIQFGIYMDD  2
FRFYLSFTLPSPGGLQPHRRFAWRVLPQGFINSPALFERALQEPLRQVSAAFSQSLLVSYMDD  3
 * *  **  *     *  *     ****   **          *              ****

(Note that by chance about 1 out of 20 amino acids would match between two random sequences, and about 1 out of 8000 would match among four random sequences.)

Over this portion, HIV and SIV are 72% similar, VLV is 56-59% similar to SIV and HIV, and BLV is 32-37% similar to the other viruses. Also note that there is no way to recombine the BLV and VLV sequences to form the HIV sequence.

(You don't have to trust my choice of which parts of the sequence to look at; you can examine the whole sequences yourself.)

Therefore, you can see from the protein sequences that HIV is not formed by combining BLV and VLV, which is strong evidence against Dr. Strecker's theory. You can also see that SIV is closer to HIV than either BLV or VLV. This supports the hypothesis that HIV and SIV are closely related.

Another smoking gun against the visna theory is that HIV-1 and HIV-2 have a vpr gene, SIV has a vpr gene, but visna and the other viruses do not. If HIV came from visna and HTLV-I, where did the vpr gene come from?

For more information about comparing sequences, you can look at "The Emergence of Simian Human Immunodeficiency Viruses", G. Meyers, K. MacInnes, B. Korber, AIDS Research and Human Retroviruses, 8(3), 1992. They look at the sequences of many lentiviruses and determine how they probably evolved.

You can look at the genetic sequences yourself at NCBI.

• HIV-1
• HIV-2
• SIV
• BIV
• HTLV-I
• Visna