Analysing the genes expressed

Did these mysterious little colonies of cubic cells hold the answers to his questions? That’s what the researcher from Liège thought… he just had to prove it. The following stage was to observe which genes were expressed by these small cells. To do this, it was necessary to treat the pieces of hysterectomies not infected by HPV differently; these pieces were provided in abundance by the Brigham operating theatres and were plunged immediately into liquid nitrogen to prevent the degradation of the RNA. In the case of classical biopsies embedded in paraffin, they undergo heat treatment using xylene and formaldehyde, which degrade the nucleic acids. Therefore, we cannot observe the expression of the genes. Using frozen samples and a laser microdissection technique, Michaël Herfs endeavoured to isolate pure populations of junction cells, endocervical cells and ectocervical cells, from which he then extracted the messenger RNA. This was then submitted to microarray analysis which allowed the researcher to characterise the complete expression of the genes from the three cellular populations studied. “It then became obvious that these three populations were very distinct from each other. In particular, some 80 genes are expressed 2.5 times more in the junction cells compared with the others, which is huge!”

If the three cellular populations express different genes, the proteins produced by each one should also be different. They can then serve as markers to easily identify the cells thanks to specific antibodies. The researchers therefore selected a few of these proteins, mainly depending on practical considerations (easily available antibodies) to serve as a marker for the presence of these cells. That’s why, for instance, keratin 7 was chosen as a marker (and patented with a view to possibly developing a diagnostic kit.

Re-enter the HPV

“Ok, we had a new population of cells, which were expressing specific genes; we had specific antibodies to identify them, now we had to find out whether this population of cells was linked in some way to infection by the HPV responsible for cancer!” The researchers then sought to see whether cancerous cells infected by HPV expressed the same genes. They examined biopsies of grade 1, 2 and 3 dysplasias as well as established cancers, both epidermoid and glandular (adenocarcinoma). Result: all the cancers and grade 2 and 3 dysplasias expressed the same genes as the junction cells. But only 20% of the grade 1 dysplasias. What was the conclusion? Several important things! On the one hand, that the cancers and precancerous lesions assessed did indeed show that they were related to the junction cells; therefore, the source of cervical cancer does indeed lie in these cells.

On the other hand, the observations on the grade 1 dysplasias were also very important. As already mentioned, the large majority of these dysplasias regress spontaneously. And yet, 80% of them didn’t express junction cell genes… “We came up with the hypothesis that if HPV infects cells other than junction cells, the lesion will spontaneously regress after six to twelve months. However, if HPV infects a junction cell, this infection will be persistent and will evolve through the different stages of dysplasia and cancer.” Remember that the question at the outset was, for Michaël Herfs, to find a way to predict which CIN1 dysplasias would regress and which ones would become aggressive. Did he answer this question and, at the same time, open the way towards tumoral marking with a prognostic value? It is very possible but our researcher remains cautious. “The next study inevitably involves verifying the clinical interest and prognostic value of such a marker. That’s why Professor Crum asked all the pathologists at the hospital to search for our marker (we chose keratin 7 for practical reasons because we already had this antibody) each time they observed a grade 1 dysplasia.”