Human-specific GAPDH qRT-PCR is an accurate and sensitive method of xenograft metastasis quantification

Abstract

Metastasis is the primary cause of cancer-related mortality. Experimental models that accurately reflect changes in metastatic burden are essential tools for developing treatments and to gain a better understanding of disease. Murine xenograft tumor models mimic the human scenario and provide a platform for metastasis analyses. An ex vivo quantitative method, gaining favor for its ease and accuracy, is quantitative reverse-transcriptase polymerase chain reaction (qRT-PCR); however, it is currently unclear how well this method correlates with gold-standard histological analysis, and its use has required detection of overexpressed exogenous genes. We have introduced a variation of the qRT-PCR method: human-specific glyceraldehyde 3-phosphate dehydrogenase (GAPDH) qRT-PCR, which allows quantification of metastasis in xenograft models without the requirement of overexpressed exogenous genes. This makes the method easily amenable to many xenograft models without alteration of the cancer cells. We determined that the method is able to detect a few human cells within abundant mouse lung tissue. Further, the human-specific GAPDH qRT-PCR is more sensitive and correlates with histological analysis in terms of determining relative metastatic burden, suggesting that human-specific GAPDH qRT-PCR could be used as a primary method for quantification of disseminated human cells in murine xenograft models.

Graphical abstract

Figure 1

Schematic representation of workflow comparing quantification of lung metastasis by histological analysis of H&E-stained, fixed thin section versus human-specific GAPDH qRT-PCR (Top) Lung processed for histology-based quantification. I, Formalin-fixed paraffin-embedded (FFPE): majority of multi-lobed lung kept in histology cassette for 24 h in formalin and then stored in 70% ethanol, followed by dehydration/clearing/embedding. II, Sectioning by microtome cutting. III, H&E staining includes deparaffinization/rehydration/staining, followed by dehydration and coverslipping. Images were captured by 2.5× magnification and metastasis quantified in ImageJ. (Bottom) Lung processed for qRT-PCR-based quantification. I, Single-lobed left lung minced/homogenized and added to II, 1 mL of TRIzol for subsequent RNA extraction. III, cDNA synthesized; IV, qPCR performed with GAPDH human-specific primers and nonspecific mouse GAPDH primers as a reference gene. Created with BioRender (https://biorender.com/).

Figure 2

Tumor volumes and histological quantification by H&E staining of fixed thin sections demonstrate that ALDH1A3 expression increases metastasis to the lungs of orthotopically established MDA-MB-231 tumors (A) ALDH1A3 overexpression increases MDA-MB-231 tumor volume and control (n = 15; ALDH1A3 overexpression n = 11, SEM error bar; one-tailed t test performed on final volume measurement). (B) MDA-MB-436 tumor volume (n = 13, SEM error bars). (C) Light microscopy image of H&E-stained lung section from mice that had MDA-MB-231 tumors; example from ALDH1A3 overexpression tumor-bearing mouse with metastatic nodes outlined. (D) Histology-quantified MDA-MB-231 lung metastasis (median, SEM error bars; Mann-Whitney test); ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

Figure 3

Validation that the human GAPDH primers are efficient human-specific primers (A and B) The qPCR amplification curves (A) and melt curves (B) generated by the human and nonspecific mouse GAPDH primers using a 2-fold serially diluted cDNA template from RNA extracted from MDA-MB-231 cells and naive NOD/SCID mouse lung. MT, melting temperature in Celsius at the peak. (C) Standard curves were generated using the qPCR amplification from (A), and the starting quantities of the 2-fold diluted cDNA are arbitrary, with the most concentrated samples set at the default setting of 1,000,000. E, efficiency; R2, the square of the correlation (the coefficient of determination). (D) Three different concentrated MDA-MB-231 and naive NOD/SCID mouse lung RNA samples were analyzed via qRT-PCR with the human and nonspecific mouse GAPDH primers.

Figure 4

The detection limit of human-specific GAPDH qRT-PCR is 100 human cells per mouse lung lobe, and the standard curve is a linear range between 100 and 1,000,000 human cells per mouse lung lobe (A) Increasing numbers of known MDA-MB-231 cells (top) or MDA-MB-436 cells (bottom) were added to naive NOD/SCID lung lobe samples and total RNA extracted and the detection limit of human-specific GAPDH primers determined by qRT-PCR. N/D, not detected. (B) Standard curve of the number of MDA-MB-231 cells (top) or MDA-MB-436 cells (bottom) added in the naive NOD/SCID lung lobe plotted against the relative amount of human GAPDH transcript detected by qRT-PCR. The normalized human GAPDH (detected with human-specific primers) per sample was made relative to the total GAPDH (detected with nonspecific mouse primers, ΔΔCt) and made relative to the 100 MDA-MB-231 cells added to a naive mouse lung lobe sample.

Figure 5

Human-specific GAPDH qRT-PCR quantification of lung metastasis demonstrates that ALDH1A3 expression increases metastasis to the lungs of orthotopically established MDA-MB-231 tumors (A) The number of MDA-MB-231 cells per lung lobe in the experimental samples is determined by using the standard curve generated in Figure 4B. Furthermore, the normalized human GAPDH (detected with human-specific primers) per sample was made relative to the total GAPDH (detected with nonspecific mouse primers, ΔΔCt). (B) The number of MDA-MB-231 cells/lung lobe in control (n = 15) versus ALDH1A3 overexpression (n = 11) samples is compared by human-specific GAPDH qRT-PCR median, SEM error bars; Mann-Whitney test; ∗p < 0.05; determined using the standard curve of known MDA-MB-231 cells/lung lobe. (C) The number of MDA-MB-436 cells/lung lobe in samples is determined by human-specific GAPDH qRT-PCR using the standard curve of known MDA-MB-436 cells/lung lobe. The human GAPDH per each sample was made relative to the total GAPDH detected by nonspecific mouse GAPDH primers.

Figure 6

Human-specific GAPDH qRT-PCR correlates with and is more sensitive than histological quantification of lung metastasis by H&E staining of fixed thin sections (A) Summary of the lung metastasis quantified in all of the experimental samples determined by human-specific GAPDH qRT-PCR (left, number of MDA-MB-231 cells/lung lobe) versus histological quantification by H&E staining of thin sections (right, percentage of MDA-MB-231 metastasis area averaged over 4 lobes of lung tissue). (B) The MDA-MB-231 lung metastasis values determined in every experimental sample by human-specific GAPDH qRT-PCR (x axis) versus histological quantification of H&E-stained, fixed thin sections (y axis) are plotted, and the Pearson correlation (R) is calculated.