Semiautomated pipeline for quantitative analysis of heart histopathology

Abstract

Background: Heart diseases are among the leading causes of death worldwide, many of which lead to pathological cardiomyocyte hypertrophy and capillary rarefaction in both patients and animal models, the quantification of which is both technically challenging and highly time-consuming. Here we developed a semiautomated pipeline for quantification of the size of cardiomyocytes and capillary density in cardiac histology, termed HeartJ, by generating macros in ImageJ, a broadly used, open-source, Java-based software.

Methods: We have used modified Gomori silver staining, which is easy to perform and digitize in high throughput, or Fluorescein-labeled lectin staining. The latter can be easily combined with other stainings, allowing additional quantitative analysis on the same section, e.g., the size of cardiomyocyte nuclei, capillary density, or single-cardiomyocyte protein expression. We validated the pipeline in a mouse model of cardiac hypertrophy induced by transverse aortic constriction, and in autopsy samples of patients with and without aortic stenosis.

Results: In both animals and humans, HeartJ-based histology quantification revealed significant hypertrophy of cardiomyocytes reflecting other parameters of hypertrophy and rarefaction of microvasculature and enabling the analysis of protein expression in individual cardiomyocytes. The analysis also revealed that murine and human cardiomyocytes had similar diameters in health and extent of hypertrophy in disease confirming the translatability of our murine cardiac hypertrophy model. HeartJ enables a rapid analysis that would not be feasible by manual methods. The pipeline has little hardware requirements and is freely available.

Conclusions: In summary, our analysis pipeline can facilitate effective and objective quantitative histological analyses in preclinical and clinical heart samples.

Keywords: Heart disease; Heart histology; Heart histopathology; Hypertrophy of cardiomyocytes; ImageJ.

Fig. 1

Overview of the analysis pipeline HeartJ. Gomori silver staining (Bright-field) or immunofluorescence staining (WGA (green), DAPI (blue), CD31 (red)) images can be acquired by microscopes or scanners. Macros created for different types of images can analyze the size of cardiomyocytes, nuclei of cardiomyocytes, and capillaries. In addition, the relationship between capillaries and cardiomyocytes can be analyzed, as well as the expression of an intracardiomyocytic signal. The results are exported as an Excel file, and an overview of the analyzed image is also created, enabling manual correction of the results

Fig. 2

Processing of Gomori silver staining in ImageJ. First, segmentation is performed using watershed segmentation. The user defines the tolerance and can add missing lines if necessary. Cardiomyocytes (CM) are identified by size and color in all instances. Capillaries are identified by their position between cardiomyocytes and size. Nuclei of cardiomyocytes are identified by size and location within cardiomyocytes. The ratio of cardiomyocytes to capillaries is evaluated. An overview image with all evaluated instances is created

Fig. 3

Processing of WGA+CD31+DAPI in ImageJ. Merge images are split into single channels first. The WGA channel is segmented using watershed segmentation. The user defines the tolerance and can add missing lines if necessary. The cardiomyocytes (CM) are identified by their size. Nuclei of cardiomyocytes and capillaries are identified in their appropriate channel by auto-threshold. By using the instances of the cardiomyocytes noncardiomyocyte nuclei can be excluded and the relationship of capillaries per cardiomyocyte can be evaluated. An overview image with all evaluated instances is created

Fig. 4

Evaluation of TAC experiment with Gomori silver staining. A Representative images of cross sections from the inner wall of the left ventricle, B of cross sections from the inner wall of the right ventricle and C of longitudinal sections from the mid-wall of the left ventricle of sham and TAC mice, Gomori silver staining (scale = 20 µm). D Cardiomyocyte area and MinFeret of sham mice (n = 9) and TAC mice (n = 9) from three cross section images per animal of the inner wall of the left ventricle with manual correction as mean value per animal, E as violin diagram of single cardiomyocytes of sham mice (n = 7401) and TAC mice (n = 6069), Median (horizontal line), 25th and 75th centiles (dotted line), and F Standard deviation (SD) of single cardiomyocytes per animal. G Cardiomyocyte area and MinFeret of sham mice (n = 9) and TAC mice (n = 9) from three cross section images per animal of the inner wall of the left ventricle without manual correction, H from three cross section images per animal of the inner wall of the left ventricle from the second batch with manual correction and I from three cross section images per animal of the outer wall of the left ventricle with manual correction and J the inner wall of the right ventricle and with manual correction. K Cardiomyocyte area and MinFeret of sham mice (n = 9) and TAC mice (n = 9) from three longitudinal section images per animal of the mid-wall of the left ventricle with manual correction. L Cardiomyocyte nuclei area and MinFeret of sham mice (n = 9) and TAC mice (n = 9) from three cross section images per animal of the inner wall of the left ventricle with manual correction

Fig. 5

Evaluation of human cohort with Gomori silver staining. A Representative images of cross sections of no aortic stenosis (AS) and AS (scale = 20 µm). B Cardiomyocyte area and MinFeret of no AS (n = 14) and AS (n = 7) from five cross section images per case without manual correction C and as violin diagram of single cardiomyocytes of no AS (n = 60 733) and AS (n = 21 644). Median (horizontal line), 25th and 75th centiles (dotted line). D Standard deviation (SD) of cardiomyocyte area and MinFeret of no AS (n = 14) and AS (n = 7) from five cross section images per patient without manual correction

Fig. 6

TAC experiment capillary density and pro-ANP expression. A Representative images of cross sections of sham and TAC mice, WGA (green), CD31 (red), and DAPI (blue) (scale = 20 µm). B Cardiomyocyte area and MinFeret of sham mice (n = 9) and TAC mice (n = 9) from three cross section images per animal of the inner wall of the left ventricle with manual correction C and their nuclei area and MinFeret with manual correction. D Capillary area and MinFeret of sham mice (n = 9) and TAC mice (n = 9) from three cross section images per animal of the inner wall of the left ventricle with manual correction. E Capillary contacts (CC) of one cardiomyocyte and ratio from CC to the area of the cardiomyocyte (CM). F Representative images of cross sections of sham and TAC mice, WGA (green), pro-ANP (red), and DAPI (blue). (scale = 20 µm). G Pro-ANP expression of sham mice (n = 9) and TAC mice (n = 9) per area of cardiomyocyte (CM) in percent (%) from three cross section images per animal of the inner wall of the left ventricle with manual correction. H Spearman correlation matrix of cardiomyocyte (CM) area, cardiomyocyte MinFeret, cardiomyocyte nuclei area, capillary contacts, Ratio of capillary contacts (CC) and cardiomyocyte area, and pro-ANP expression in percent of cardiomyocyte area