Repeated assessment of changes in testes size in canaries by X-ray computer tomography

Abstract

Numerous studies have evaluated changes in time of testicular development in birds by exploratory laparotomy or post-mortem autopsy. The invasive nature of these approaches has obviously limited the frequency at which these measures can be collected. We demonstrate here that accurate assessment of gonadal size can be reliably and repeatedly obtained by computer-assisted X ray tomography (CT scans). This approach provides images of the testes in the three orthogonal planes that allow measuring either the largest diameter or even the volume of the testes, providing results that match those obtained by surgical approaches.

1. Introduction

Most avian species, in particular those living in the temperate zone, display marked seasonal changes in gonadal physiology that have been the topic of intense research during the last century since the seminal paper of Rowan (Murton and Westwood, 1978; Nicholls et al., 1988; Rowan, 1925; Williams, 2012). For example, testis mass changes in European starlings (Sturnus vulgaris) from about two hundred milligrams when birds are held in short days to over one gram after birds have been exposed for a few weeks to a long photoperiod mimicking summer long days (Young et al., 2001)(Reviewed in (Nicholls et al., 1988)). In canaries, the testes diameter increases from around 0.5 mm in short day males to more than 5 mm after photostimulation (Hurley et al., 2008). Similar changes with even a larger amplitude, presumably as a result of selection for active reproduction, are also observed in Japanese quail (Coturnix japonica) with testes mass going from a few milligrams in short days up to several grams in long days (Follett, 1976; Nicholls et al., 1973)(up to over 10 grams for paired testes weight in our quail colony; J. Balthazart personal observation). Such seasonal plasticity in testes size is absent in many mammals including the widely studied rodent species that serve as model systems (i.e., mice and rats) but is also observed, although with a smaller amplitude, in some mammalian species that display photoperiodic responses such as Dungarian hamsters (Phodopus sungorus) (Niklowitz et al., 1994). Correlatively, an intense apoptosis activity has been detected during the period of gonadal regression in Sertoli cells of the starling testes while this phenomenon does not seem to occur in the mammalian testis (Young et al., 2001).

Historically changes in testicular size have been assessed either by autopsy when birds were killed at the end of an experiment (different groups of subjects then have to be used to assess different physiological conditions) or by repeated laparotomy, i.e. opening of the abdominal cavity in anesthetized subjects followed by measurement of the testes in situ. This latter procedure is very well tolerated by some species of birds such as Japanese quail or ring doves (Streptopelia risoria) in which we were for example able to follow testes and ovary changes during six to nine months by laparotomies repeated every four weeks (Balthazart et al., 1981; Cheng et al., 1980). However other species tolerate this type of surgery less easily: some are even susceptible to die under anesthesia and others take several hours or days to recover before they again display a normal level of behavioral activity.

The recent development of biomedical imaging techniques that can visualize internal structures in a non-invasive manner offers in theory an alternative procedure to these surgical approaches. For example, magnetic resonance imaging (MRI) has been employed in this context to visualize and measure testes and ovarian follicles in garden warblers (Sylvia borin) of both sexes. This procedure provided high resolution images of the gonads but the imaging procedure of a given bird took about 20 min (Czisch et al., 2001). The equipment used is also very expensive and requires a specialized (technical) engineer for its operation.

X-ray based radiography is commonly used as a diagnostic tool in veterinary medicine (Latimer, 2019)(see https://www.msdvetmanual.com/) and has been used in the past to evaluate the size of internal organs in small birds (Beregi et al., 1999). Radiographic images are however complex and difficult to interpret since they represent three-dimensional objects in only two dimensions. This limitation is overcome by computer-assisted tomography (CT) scanning procedures that provide 3D-views of internal structures. We illustrate here how this method can be used to visualize and repeatedly quantify changes in gonadal size in a small songbird species, the canary (Serinus canaria).

2. Material and Methods

2.1. Subjects

The present experiments were carried out with canaries of the Fife Fancy breed that have been shown to have retained a photoperiodic response and thus exhibit seasonal changes in testes volume as a function of changes in the ambient photoperiod (Cornez et al., 2020). All birds were purchased from a local supplier in Belgium and kept until used for experiments under a short day photoperiod (8L:16D) in cages of approximately 6 subjects or in a larger aviary containing up to 50 subjects. All birds received food (canary seeds) and water ad libitum during all experiments. Anise scented sand, cuttlebones and perches were always present in the cages as enrichments. Additionally, the birds received egg food once a week. All experimental procedures complied with Belgian laws concerning the Protection and Welfare of Animals and the Protection of Experimental Animals, and experimental protocols were approved by the Ethics Committee for the Use of Animals at the University of Liège (Protocol 1739).

2.2. Imaging procedure

Before imaging, birds were first fasted for two-three hours since we observed that food in the digestive tract caused artifacts in the images obtained by CT scan. Birds were then anesthetized with Isoflurane (3% for induction, 1.5% for maintenance) and placed on the platform of the CT scanner with their head inserted into a 10 ml syringe connected to the Isoflurane supply. We used a small animal irradiator and scanner (Small Animal Radiation Therapy with advanced precision, SmART) instrument from Precision X-Ray (North Branford, CT 06471 USA) designed to image, target and irradiate cells and small animals up to the size of rats (see https://precisionxray.com/small-animal-igrt/). This scanner provides images that have a 0.1 mm resolution.

A pre-scan of the whole body was initially carried out to locate more precisely the volume to be investigated and this volume was then defined in all three dimensions by moving cursors on the computer screen with the mouse. The acquisition scan was then started. The scanning parameters were 40 Kv, 8mA and a voxel resolution of 0.1x0.1x0.1 mm.

If all birds were roughly positioned in the same way, there was no need to repeat the pre-scan for each subject and the scanning of one bird then only took a few minutes. Different subjects could thus be quickly scanned in sequence and the entire procedure including induction of the anesthesia, positioning in the scanner and acquisition of the images took less than 5 min, allowing us to image 10 to 12 birds in one hour. Birds were then placed in a recovery cage lightly heated by a lamp; they recovered and were perching within minutes.

2.3. Image analysis

The scanning process generates a folder containing about 400 images coded in DICOM format (Digital Imaging and Communication in Medicine, suffix .dcm). These can easily be analyzed with any DICOM Viewer and Image J, a free software produced and maintained by the National Institutes of Health USA (NIH, Bethesda; https://imagej.nih.gov/) does a great job in this task.

The different steps in these analyses start with the IMPORT of an IMAGE SEQUENCE obtained by selecting all the .dcm files located in the folder resulting from one CT scan. Using various options in the IMAGE menu, one can then ADJUST the brightness and contrast of the images and eventually turn them into pseudo-colors by using the LOOKUP TABLES menu. This often makes it easier to distinguish individual structures in the body since the human eye is more sensitive to changes in color than changes in the gray scale level. In the IMAGE menu, STACKS sub-menu, one can then click on ORTHOGONAL VIEWS to obtain a vision of the objects in sagittal and horizontal view in addition to the coronal view initially displayed. In the same sub-menu, clicking on 3D PROJECT will generate a three-dimensional reconstruction of the bird body that can be rotated and viewed in any angle. Moving between these different sets of images allows detection of the structures of interest, in this case the testes. By moving with the mouse the intersection of the two thin lines indicating the level of viewing one can then navigate through the entire volume that has been acquired.

Accurate measurements can then be obtained with the use of the SET SCALE function in the CALIBRATE menu. This will calibrate the pixels (voxels) of images into millimeters. This requires one to have imaged an object of known length at the same time as the bird. Practically, all our birds are wearing a leg band of 4 mm in length and we used this object to calibrate the images. Once this is done, drawing a line across the largest diameter of each testis in the coronal plane (MEASURE in the ANALYZE menu) will provide a measure in mm that will be added into a data table. This largest diameter can be reported as done previously by Hurley and collaborators (Hurley et al., 2008). One can additionally measure the length of the testis in the rostro-caudal axis (sagittal or horizontal plane) and compute testis volume with the formula V=4/3 π a² b, where a is the half width and b is the half length (long axis) (see (Bentley et al., 2003)).

3. Results

Figure 1 illustrates the results on one CT scan obtained in a male canary with more or less fully developed testes as viewed in the three orthogonal planes and as a 3D reconstruction. As can be seen by scrolling through the different levels of the images, it is easily feasible to identify and then measure the testis size. The largest diameter in this bird was 6.3 mm and its length 7.5 mm giving an estimated volume of 155.8 mm³.

Fig 1.

CT scans of the testes in a reproductively active male canary. The different panels illustrate the data as acquired in coronal orientation and then their presentation in the sagittal and horizontal plane, as well as the 3D-reconstruction of the entire body. The arrows in the 3D panel highlight the boundaries of one testis visible in superposition of one leg. The double-headed lines in the three other panels show the measures of the largest diameter of one of the testes in the coronal plane and then its antero-posterior length in the sagittal and horizontal planes. The thin blue lines can be moved in the coronal representation by the program to select the levels displayed in the other panels. Ant.: anterior, Post.: posterior.

These images could be collected repeatedly in the same subjects. Figure 2 illustrates the testes of the same male imaged on four successive occasions during a 14-week period. The bird had been kept on short days (8 hours of light: 16 hours of darkness per day or 8L:16D) for at least 6 months during which it became photosensitive and developed nearly maximal testes size after dissipation of photorefractoriness had occurred, as normally observed in this breed of canaries (see Cornez et al., 2020; Balthazart Jacques and Chiver Ioana, unpublished data) . On week 0 it was, after a first scan, transferred into a long day photoperiod (18L:6D) that induced a slight increase in testes size as attested by the scans performed on weeks 4 and 10. The bird was then returned to short days and its testes collapsed within 4 weeks revealing that the bird had developed a relative photorefractoriness (Dawson et al., 2001). Note that the bird was simply anesthetized and placed on the holding plate of the scanner but did not have to be identically positioned each time since the software allows sectioning through the volume to identify the largest testis diameter irrespective of the bird position during image acquisition.

Fig. 2.

Coronal views through the largest diameter of the testes in a male that was kept in short days (SD) and photosensitive (week 0), transferred into long days (LD) and scanned after 4 and 10 weeks and then returned to SD and scanned 4 weeks later. Note that at this last time point testes were very small and no longer in the same plane so that only one of them can be seen in the same coronal section (arrows). The bottom panel illustrates the mean of the diameter of the two testes at these different time points.

Some of the subjects used in these experiments were euthanized to collect their brain for independent studies not presented here. This allowed us to compare images obtained by CT scans with the actual size of the testes as observed by dissection. One example is presented in figure 3. As can be observed, there is an excellent match between CT scan images and the actual size of the testes. This is illustrated by the fact that the left testis had a slightly elongated (oblong) shape while the right testis appeared as an almost perfect sphere. This difference is clearly observed in the scans.

Fig. 3.

Comparison of the testes views obtained by CT scans and by dissection after euthanasia. Note that the largest extension of the left (L) and right (R) testis cannot be visualized in the same sagittal or horizontal plane and two images are therefore presented.

Finally, we confirmed this concordance between measures obtained by CT scan and dissection by calculating their correlation in a batch of 15 birds that had been scanned and killed 5 days later in order to collect their brain for an experiment not described here. The correlation coefficient between the mean testis length measured in scan images and during dissection was equal to 0.890 (R²=0.792), which is highly significant (p<0.0001). The regression line of these two sets of data was Y=0.914 X + 0.162 (with Y = dissection and X= CT scan), indicating a slope close to 1 and an intercept close to the axes origin (see Fig 4); the two sets of measures thus provide very similar size estimates. We also correlated the mean volume of the two testes measured by CT scan and the mean weight of these testes measured after dissection and this correlation was similarly high and significant (r=0.922, r²=0.851, Y=0.9703 X - 3.285; p<0.0001).

Fig. 4.

Correlation between the mean length of the two testes as measured in CT Scans and by direct observation after dissection in a group of 15 male canaries.

4. Discussion

We demonstrate here that X-ray computer-assisted tomography allows visualizing and measuring in a rapid and accurate manner the testes in small passerine birds such as canaries. The values of largest testis diameter or testis volume obtained in this way are similar to those obtained by laparotomy or post-mortem autopsy as reported previously (Bentley et al., 2003; Hurley et al., 2008). The procedure is rapid (less than 5 min per bird) and non-invasive. We elected to image the birds under light isoflurane anesthesia to decrease stress, facilitate manipulations and ensure that subjects would not move during the scanning but restraining the bird and placing it in a dorsal decubitus position inducing tonic immobility (Jones, 2019) would clearly work in most cases.

The quality and resolution of images obtained by magnetic resonance imaging (MRI) as illustrated in Czisch et al., (2001) are admittedly a bit better especially for soft tissues but this approach is slower and most importantly more expensive. The use of Micro CT scanners is probably 3 to 10 times cheaper than the use of MRI (https://www.monash.edu/researchinfrastructure/mbi/forms-and-policies/prices). CT scanners are also probably more broadly available than MRI equipment although they are obviously not usually present in research labs interested in avian biology. These scanners however exist in substantial numbers in biomedical laboratories and collaborations can be established to get access to the equipment for short periods that are sufficient to monitor substantial numbers of subjects. Learning to manipulate the software driving the scanner is relatively straightforward and therefore specialized engineers are not needed to operate the equipment, which is clearly the case for MRI.

This approach permits following accurately the changes in testes volume in small birds that would be adversely affected by laparotomies, as was done previously by MRI in Pied flycatchers, Ficedula hypoleuca (Coppack et al., 2008). Since the approach is not invasive contrary to surgical laparotomy, it allows repeating measures as frequently as desired. These estimates of gonadal state obtained with an increased temporal resolution via non-invasive methods will provide an ability to better assess “endpoint” changes in reproductive physiology induced by changing environmental cues and endogenous rhythms and integrate them with studies investigating endocrine, neuroendocrine, neural, or genomic mechanisms. This method will also provide an impetus for additional comparative studies in species that may otherwise have been neglected because of limitations in using invasive methods such as laparotomy. This will permit the development of experimental protocols requiring these multiple assessment of changes in gonadal morphology.

Highlights.

-Computer tomography (CT) scanner can easily visualize testes in songbirds

-Changes in testes volume can in this way be measured in a non-invasive manner

-A full body scan takes less than 5 minutes

-Measures can be frequently repeated since they induce very little stress

-CT scanning is cheaper, easier and faster than Magnetic resonance imaging