Chapter 6:  Mapping human genes by using human–rodent somatic cell hybrids

Assigning genes to chromosomes

The technique of somatic cell hybridization is extensively used in human genome mapping, but it can in principle be used in many different animal systems. The procedure uses cells growing in culture. A virus called the Sendai virus has a useful property that makes the mapping technique possible. Each Sendai virus has several points of attachment, so it can simultaneously attach to two different cells if they happen to be close together. However, a virus is very small in comparison with a cell, so the two cells to which the virus is attached are held very close together indeed. In fact, the membranes of the two cells may fuse together and the two cells become one—a binucleate heterokaryon.

If suspensions of human and mouse cells are mixed together in the presence of Sendai virus that has been inactivated by ultraviolet light, the virus can mediate fusion of the cells from the different species (Figure 6-23). When the cells have fused, the nuclei subsequently fuse to form a uni-nucleate cell line composed of both human and mouse chromosome sets. Because the mouse and human chromosomes are recognizably different in number and shape, the two sets in the hybrid cells can be readily distinguished. However, in the course of subsequent cell divisions, for unknown reasons the human chromosomes are gradually eliminated from the hybrid at random. Perhaps this process is analogous to haploidization in the fungus Aspergillus.

The loss of human chromosomes can be arrested in the following way to encourage the formation of a stable partial hybrid. The cells used are mutant for some biochemical function; so, if the cells are to grow, the missing function must be supplied by the other genome. This selective technique results in the maintenance of hybrid cells that have a complete set of mouse chromosomes and a small number of human chromosomes, which vary in number and type from hybrid to hybrid but which always include the human chromosome carrying the wild-type allele defective in the mouse genome.

Let’s look at the specific genes that make the selective system work. In cells, DNA can be made either de novo (“from scratch”) or through a salvage pathway that uses molecular skeletons already available. The selective technique involves the application of a chemical, aminopterin, that blocks the de novo synthetic pathway, confining DNA synthesis to the salvage pathway. Two essential salvage enzymes, thymidine kinase (TK) and hypoxanthine-guanine phosphoribosyl transferase (HGPRT), are relevant to the system, as shown in the following two reactions:

The mouse cell line to be fused is genetically unable to make TK because it is homozygous for the allele tk⁻, whereas the human cell line is genetically unable to make HGPRT because it is homozygous at another locus for the allele hgprt⁻. So the genotypes of the two fusing cell lines are:

Because each is deficient for one enzyme, neither the mouse nor the human cells are able to make DNA individually. In the hybrid cells, however, the tk⁺ allele complements the hgprt⁺ allele, so the cells can make both enzymes. Therefore, DNA is synthesized and the cells can proliferate. Most human chromosomes are eliminated from the hybrid cell cultures because their loss has no effect on the cultures’ ability to grow. But, to continue to grow in medium containing hypoxanthine, aminopterin, and thymidine (HAT medium), a hybrid culture must retain at least one of the human chromosomes that carries the tk⁺ allele.

Luckily, the progressive elimination of the human chromosomes from the fused cell lines can be followed under the microscope because mouse chromosomes can easily be distinguished from human chromosomes. Moreover, chromosome stains such as quinacrine and Giemsa reveal a pattern of banding within the chromosomes. The size and the position of these bands vary from chromosome to chromosome, but the banding patterns are highly specific and invariant for each chromosome. Thus, it is easy to identify the human chromosomes that are present in any hybrid cell (Figure 6-24). Different hybrid cells are grown separately into lines; eventually a bank of lines is produced that contains, in total, all the human chromosomes.

With a complete bank of chromosomes, we can begin to assign genes or markers to chromosomes. If the human chromosome set is homozygous for a human molecular marker—such as an allele that controls a cell-surface antigen, drug resistance, a nutritional requirement, a specific protein, or a DNA marker—then the presence or absence of this genetic marker in each line of hybrid cells can be correlated with the presence or absence of certain human chromosomes in each line. Data of this sort are presented in Table 6-2 in which “+” means presence and “−” means absence of the genetic marker. We can see that, in the different hybrid cell lines, genetic markers 1 and 3 are always present or absent together. We can conclude, then, that they are linked. Furthermore, the presence or absence of genes 1 and 3 is correlated with the presence or absence of chromosome 2, so we can assume that these genes are located on chromosome 2. By the same reasoning, gene 2 must be on chromosome 1, but the location of gene 4 cannot be assigned. Large numbers of human genes have now been localized to specific chromosomes in this way.

Chromosome mapping

In the preceding subsection, we discussed the use of human-rodent cell hybrids to assign genes to chromosomes. This technique can be extended to obtain mapping data. One extension of the hybrid cell technique is called chromosome-mediated gene transfer. First, samples of individual human chromosomes are isolated by fluorescence-activated chromosome sorting (FACS) (Figure 6-25). In this procedure metaphase chromosomes are stained with two dyes, one of which binds to AT-rich regions, and the other to GC-rich regions. Cells are disrupted to liberate whole chromosomes into liquid suspension. This suspension is converted into a spray in which the concentration of chromosomes is such that each spray droplet contains one chromosome. The spray passes through laser beams tuned to excite the fluorescence. Each chromosome produces its own characteristic fluorescence signal, which is recognized electronically, and two deflector plates direct the droplets containing the specific chromosome needed into a collection tube.

Then a sample of one specific chromosome under study is added to rodent cells. The human chromosomes are engulfed by the rodent cells and whole chromosomes or fragments become incorporated into the rodent nucleus. Correlations are then made between the human fragments present and human markers. The closer two human markers are on a chromosome, the more often they are transferred together. Some results of these kinds of mapping are shown in Figure 6-26.

A second method, called irradiation and fusion gene transfer (IFGT), an extension of the chromosomemediated gene transfer, is designed to generate a higherresolution map of molecular markers along a chromosome. The procedure is to irradiate human cells with 3000 rads of X rays to fragment the chromosomes, then fuse the irradiated cells with the rodent cells to form a panel of different hybrids. In this case the hybrids have an assortment of fragments of human chromosomes, as diagrammed in Figure 6-27. Most of the fragments are seen to be embedded into the rodent chromosomes, but truncated human chromosomes also can be found. First, the retention of various molecular markers in the hybrids is calculated. The next step is to calculate the frequency of coretention of pairs of human molecular markers. The assumption is made that closely linked markers will be coretained at high frequencies because there is a low probability that a radiation-induced break will occur between the loci. Distant markers and markers on different chromosomes should be retained at frequencies close to the product of individual retention frequencies. A mapping unit cR₃₀₀₀ is calculated, which can be calibrated to approximately 0.1 cM (m.u.).

One of the advantages of this method is in sample size; a large number of radiation hybrids can be amassed relatively easily. A standard panel of only 100 to 200 hybrids is enough to generate a high-resolution “cR₃₀₀₀ map” of the human genome, with a tenfold better resolution than a map based on RF.