Early embryo elongation involves coordinated cellular and tissue behaviors that are readily observable in the chick embryo vertebrate model system. Easily identifiable morphological landmarks are crucial to obtain reliable morphometric data, particularly when assessing tissue elongation over time. The posterior end of the primitive streak marks the caudal end of the chick embryonic tissue. However, the identification of its precise location is ambiguous, especially to the untrained eye. Herein, we assessed if the posterior limit of the area pellucida (pPL), which is readily recognizable due to the optical contrast with the area opaca, is a valid proxy for the caudal limit of the primitive streak. Measurements of total embryo length were performed in multiple images of chick embryos over time using both caudal landmarks. We found that the pPL offered greater precision and a higher degree of inter-user reproducibility, when compared to the end of the primitive streak. Importantly, our work uncovers a quantitative proportionality between embryo length measurements using the end of the primitive streak and the pPL as caudal landmarks. We have thus validated the pPL as a reliable morphological proxy for the end of the primitive streak in chick embryo elongation studies.

Vertebrate embryo body elongation requires the orchestrated growth of multiple tissues, involving cell proliferation, migration and rearrangements along the anterior-posterior axis^[1]. During early development, the embryo body elongates as epiblast cells, migrate through the primitive streak and are specified into multiple tissues. Live-imaging studies have been used with great success to characterize the dynamics of these processes, as well as to evaluate the impact of different culture conditions or drug exposure on embryo elongation^[1]. The chick embryo is an ideal model for such experimental approaches, due to the availability of in ovo and ex-ovo culture systems amenable to live imaging approaches, and also because it presents striking morphological similarities to human development in early stages^[2].

The choice of specific morphological landmarks is a critical step when characterizing embryo elongation. These should be easily and reproducibly identifiable throughout the whole course of the experiment, and also among distinct users. When studying net embryo elongation at the expense of the epiblast tissue, it is useful to measure embryo length until the end of the primitive streak. However, it is not trivial to pinpoint this morphological landmark in a reproducible manner, even when analyzing the same embryo over time. This introduces a degree of inaccuracy to the measurements performed, and strongly hinders automated image recognition and analysis. The posterior limit of the area pellucida (henceforth designated pPL) is located immediately caudal to the end of the primitive streak. The pPL is readily recognizable due to the optical contrast between the area pellucida and the area opaca, and has been used to characterize embryo elongation rates in several studies^[3][4][5][6]. However, to what extent the pPL can be used as a reliable proxy for the end of the primitive streak has not yet been assessed.

The objective of this work was to validate the easily identifiable posterior limit of the area pellucida (pPL) as a reliable proxy for the end of the primitive streak for chick embryo elongation studies.

We performed time-lapse imaging of chick embryos cultured in New^[7] or EC^[8] systems (n=3 each) from early gastrulation to somitogenesis stages (HH4 to HH10^[9]) (Fig. 1A). Embryo length was measured from the crown to the end of the primitive streak (C-PS; Fig. 1B, C) and to the pPL (C-pPL; Fig. 1B, D). Measurements were performed in multiple independent frames per developmental stage (n=5 each) for each embryo (total n=6). We found that embryo growth over time displays a similar trend independently of the morphological landmark used (Fig. 1C, D). In fact, both C-PS and C-pPL measurements increase steadily in the embryonic stages analyzed.

To test the reproducibility of the results when the measurements are performed using each landmark, C-PS and C-pPL were measured in a single embryo over time by 3 operators with different years of experience with the chick embryo model (Fig. 1E, F). Figure 1E evidences significant variability in C-PS length measured by distinct users, which can differ up to 0.84 mm (HH5; 2.81 mm average length). This stems from differences in where each operator positions the end of the primitive streak. Importantly, the variability of the C-pPL measurement performed by the independent users is almost negligible (Fig. 1F). This highlights that the contrast between the posterior border of the area pellucida and the area opaca is an easily recognizable landmark for both experienced and inexperienced experimentalists. The pPL landmark thus allows for high precision measurements for embryo length studies.

For pPL to be a valid proxy for the end of the primitive streak, the distance between the two landmarks (PS-pPL) should be approximately constant over time; i.e., any variation in C-pPL length should be exclusively due to a respective variation in C-PS. We found that this is the case, since the natural variations observed in PS-pPL distances over time (Fig. 1G) were neglectable when compared to the progressive increase in C-pPL length (Fig. 1D). These findings evidence that C-pPL variation over time reliably represents C-PS dynamics, thus validating the use of the pPL as a precise and reliable morphological landmark for chick embryo elongation studies.

Finally, when C-PS values were plotted against their respective C-pPL, for all embryos and developmental stages analyzed, a direct proportionality was found represented by the equation [C-PS] = -0.378 + 0.947[C-pPL] with R2 = 0.949 (mean error of estimated C-PS = 0.206±0.132). This means that the absolute value of C-PS length can be directly calculated from C-pPL measurement.

Altogether, our results evidence that the posterior limit of the area pellucida (pPL) is a reliable morphological proxy for the end of the primitive streak. The pPL landmark provides higher precision measurements and minimizes inter-user variability. Importantly, we describe a novel quantitative relationship between C-PS and C-pPL, which represents a powerful tool to directly infer C-PS from experimentally-measured C-pPL.

This study was performed in HH4-HH10^[9] embryonic stages, thus the pPL landmark was only validated for this developmental time window. Although the findings are consistent in both New and EC culture systems, the sample size is limited and may not reflect the variability that could be found with a larger population set.

With the recent technological developments, there is an increased usage of live-imaging approaches to study embryo elongation and elucidate the mechanisms underlying this highly coordinated event. This may involve experimentally challenging the system by altering gene expression, signaling pathway modulation, microsurgical approaches, among others. As such, a thorough and precise characterization of chick embryo elongation in control conditions is more relevant than ever. The pPL herein validated is a powerful landmark for such studies, as it provides greater precision and inter-user reproducibility in embryo length measurements. Finally, the clear optical contrast between the area pellucida and area opaca may allow the precise identification of the pPL landmark by automated image recognition and analysis tools.

Chicken embryos and culture

Fertilized chicken (Gallus gallus) eggs were provided by commercial sources and incubated in a humidified atmosphere and controlled temperature (38ºC). HH4-5^[9] embryos were cultured in either New^[7]or EC^[8] culture systems for up to 24 h at 38ºC in a humidified atmosphere.

Image acquisition and analysis

Time-lapse movies were performed using a Zeiss SteREOLumarV12 Stereomicroscope coupled with a Zeiss Axiocam MRc camera. Measurements were performed in 5 independent frames per developmental stage for each embryo analyzed (150 images in total), using Axiovision Se 64 Rel 4.9.1 (Carl Zeiss) or FIJI^[10] software. Data analysis was performed using Excel and SPSS software.

This work was supported by FCT, Portugal (grant PTDC/BEX-BID/5410/2014) and Research Center Grant UID/BIM/04773/2013 CBMR 1334. The Light Microscopy Unit was partially funded by PPBI-POCI-01-0145-FEDER-022122. TPA and ACMF were supported by FCT fellowships SFRH/BD/84825/2012 and PTDC/BEX-BID/5410/2014, respectively.

We thank Gil Carraco for help in time-lapse imaging and Ana P. Martins-Jesus for performing measurements as an independent user. We thank Gil Carraco, Ana P. Martins-Jesus, Isabel Duarte, Ramiro Magno and Paulo Martel for helpful discussions. We acknowledge the staff of the Light Microscopy Unit of CBMR-UAlg for technical support. ACMF thanks José and Ana Paula Fernandes for their constant support in her growth.

The authors declare no conflicts of interest.

Not Applicable.

No fraudulence is committed in performing these experiments or during processing of the data. We understand that in the case of fraudulence, the study can be retracted by ScienceMatters.