Droplet digital PCR is the most robust method for absolute nucleic acid quantification. However, RNA is a very versatile molecule and its abundance is tissue-dependent. RNA quantification is dependent on a reference control to estimate the abundance. Additionally, in cancer, many cellular processes are deregulated which consequently affects the gene expression profiles. In this work, we performed microarray data mining of different childhood cancers and healthy controls. We selected four genes that showed no gene expression variations (PSMB6, PGGT1B, UBQLN2 and UQCR2) and four classical reference genes (ACTB, GAPDH, RPL4 and RPS18). Gene expression was validated in 40 acute lymphoblastic leukemia samples by means of droplet digital PCR. We observed that PSMB6, PGGT1B, UBQLN2 and UQCR2 were expressed ~100 times less than ACTB, GAPDH, RPL4 and RPS18. However, we observed excellent correlations among the new reference genes (p < 0.0001). We propose that PSMB6, PGGT1B, UBQLN2 and UQCR2 are housekeeping genes with low expression in childhood cancer.

Leukemia is the most common childhood malignancy in Mexico, representing more than 50% of all childhood cancers. Although treatment leads to a survival of up to 90% in developing countries, in our country, it is less than 65%. Additionally, ~30% of patients relapse with poor prognosis. Alternative splicing plays an important role in transcriptome diversity and cellular biology. This mechanism promotes an increase in the assortment of proteins with potentially distinct functions from a single gene. The proliferating cell nuclear antigen (PCNA) gene encodes two transcripts for the same protein of 261 amino acids, which is associated with several important cellular processes and with several types of cancer. However, the diversity of the transcript variants expressed in this condition is not clear. Then, we used microarray gene expression to identify changes in the exon expression level of PCNA. The data were validated using RT‒PCR and Sanger sequencing, and three additional transcripts (PCNA_V3, PCNA_V4, and PCNA_V5) were identified. Computational analyses were used to determine the potential proteins resulting, their structure, and interactions with PCNA native protein and themselves. Additionally, the PCNA transcript variants were inhibited using specific siRNA, determining that their inhibition contributes to the malignant characteristics in vitro. Finally, we quantified the PCNA transcript variants in acute lymphoblastic leukemia samples and identified their expression in this disease. Based on the clinical characteristics, we determined that PCNA_V2 and PCNA_V4 are expressed at significantly low levels in relapsed B-ALL patients. We conclude that the low expression of PCNA_V2 and PCNA_V4 could be a potential molecular marker of relapse in acute lymphoblastic leukemia patients.