Abstract

Many fishes possess epidermal club cells that are the presumptive source of chemical alarm cues and also play a role in innate immune defense. Club cell density has been estimated in some studies but a standardized method for quantifying these cells is lacking.

Here, we assess the repeatability of estimating club cell density in fathead minnows (Pimephales promelas). Thin-sectioned histological samples of fathead minnows were stained and mounted on slides and then digitally scanned for scoring. We estimated epidermal area using the segment tool in ImageJ to simulate the traditional method of scoring microscope slides using an ocular micrometer, where epidermal area was estimated by the lengths of straight-line segments of epidermal thickness and length of the tissue sample. The second approach measured epidermal area using the freehand tool in ImageJ.

The R² value for repeated estimates of club cell density (club cells/mm²) ranged from 0.959 and 0.969, depending on the method used to estimate epidermal area. Measurement error in estimates of epidermal area was greater than measurement error in cell counts and the freehand tool was more repeatable than the segment tool as a method to measure epidermal area.

We applied these methods to test differences in club cell densities between two sources of fathead minnows; wild-caught fish versus lab-reared fish provided by the Environmental Protection Agency. Lab-reared fish had higher densities of club cells than wild-caught fish did, likely reflecting differences in body condition.epidermal club cells

Sample preparation

Adult wild-caught fathead minnows collected from local prairie pothole lakes near Fargo, North Dakota, USA were obtained from a local bait dealer. Laboratory-reared fathead minnows were obtained from the U.S. Environmental Protection Agency (EPA) in Duluth, MN, USA. Fish were sacrificed using a lethal dosage of methane tricainesulfonate (~500 mg/L; IACUC protocol #A18054) and a 3-4 mm section of skin was excised from the nape from eight wild-caught minnows and 11 lab-reared minnows. Samples were fixed in 10% formalin solution for 24 h then set in parafilm for 8 h. Samples were stained with periodic acid Schiff reagent and counterstained with hematoxylin (PAS-H) and subsequently thin-sectioned and mounted on slides.

Digitization and estimation of club cell density

Slides were digitally scanned using a MoticEasyScan slide scanner using Plan Apochromatic objective (20X/0.75) with image detail equivalent to 40X lens. Club cells appear as large, unstained, circular cells within epithelial tissue (Figure 1). To assess the reliability and accuracy of estimates of club cell density, we scored the images twice using ImageJ software (Schneider et al. 2012). Images were enlarged to the optimal equilibrium between increasing visibility of structures without loss of resolution due to pixelation. We counted club cells visible along the entire length of epidermis visible in each image.

We used two methods to calculate epidermal area using ImageJ. The first method mimicked the method used in the pre-digital era (i.e., Wisenden & Smith 1997, 1998) when measurements were taken directly from a microscope slide using an ocular micrometer. We used the straight segment tool in ImageJ to measure the thickness of the epidermis in one location, typically near the center of the section, which seemed to represent typical thickness. Epidermal length was measured by summing several straight segments along the length of the sample, starting new segments as necessary to follow the curvature of the histological section (Figure 1a). The lengths of the segments were summed, then multiplied by epidermal thickness to estimate epidermal area. The second method was to use the freehand tool in ImageJ to trace along the outer edges of the epidermis and then used ImageJ to calculate the area of the enclosed area (Figure 1b). For both methods, each slide was scored twice, at separate sittings.

Data analysis

We assessed repeatability by regressing the estimates of club cell density from the second set of scores against the values obtained from the first set of scores. Perfect repeatable would show an R² of 1. Because density is a ratio, error in estimates of club cell density could come from errors in counting the number of cells (numerator) or in estimating the area of epidermal tissue (denominator). Therefore, we calculated R² values for second scores versus first scores for club cell counts and for estimates of epidermal area. We used Mann-Whitney U tests to compare club cell densities between wild-caught and lab-reared fathead minnows. The statistical package used was SPSS 28.0.