DIFFERENTIAL DISPLAY OF mRNAs FROM THE ATRIOVENTRICULAR REGION OF DEVELOPING CHICKEN HEARTS AT STAGES 15 AND 21

Department of Biological Sciences, University of Iowa, Iowa City, Iowa 52242-1324

Received 11/15/95; Accepted 11/28/95; On-line 1/1/96

Differential display of mRNAs from AV canal regions of stage 15 and stage 21 hearts
Different combinations of five (T₁₁)MN anchoring primers and three arbitrary 10-mers were used in the differential display. Of all 15 possible primer combinations, four (AP1/AR1, AP1/AR3, AP2/AR1, AP2/AR3 ) displayed reproducible and differential patterns of expression (Fig 1). The differentially expressed cDNA fragments were readily identified from either stage 15 sample ( bands a-e of Fig 1) or stage 21 sample ( bands 1-6 of Fig 1). In addition, there existed many cDNA fragments showing slightly differential expression.

To identity cardiac-specific differential expression between stage 15 and stage 21, we also displayed cDNAs prepared from RNA samples isolated from the same staged embryo tissues without hearts. Examples of such display and comparison are shown in Fig 2.

One of cDNA fragments indicated by arrow in Fig 2A could be thought as a cardiac-specific and differentially expressed gene if just compared to the differential expression patterns from heart tissue. By the addition of differential expression patterns generated from the same stage embryos without hearts( Fig 2B and 2C), such false positive clone could then be ruled out, because of the presence of an equally expressed band ( indicated by arrows in 15E and 21E of Fig 2B and 2C). Conversely, cardiac-specific but not differentially expressed genes could be readily identified by comparisons of displaying patterns from heart and embryo without heart (indicated by arrowheads in Fig 2).

From the displaying sequencing gels, we cut out over 20 different bands, ranging in size from 200bp to 500bp. These bands were subcloned and their sequences were determined. Table 1 summarizes 14 cDNA clones. From the sequence comparison to the GenBank database, 4 clones represent known genes, including protein synthesis initiation factor eIF-4AII ( clone 15A ), phospholamban (clone 15H16), skeletal alpha-tropomyosin (E13) and alpha-actin (clone H2). These cloned cDNA fragments were further used as probes in the whole-mount in situ hybridization screening. Of 12 clones tested, 10 probes show specific hybridization to the heart and only 2 probes did not detect significant signal. This is in contrast to the evaluation of cloned cDNA fragments by Northern blot analysis. The rate of success in identifying message sizes by Northern blot analysis with these cDNA probes was very low (29%). We chose 3 clones for further evaluation of their expression patterns in the AV canal regions of stage 15 and stage 21 hearts.

Phospholamban is differentially expressed in AV canal regions of developing hearts between stage 15 and stage 21
From the DNA sequence, we found that the clone 15H16 contained 320bp upstream from the poly(A) tail of the chicken phospholamban transcript (24). Phospholamban is a protein regulating the sarcoplasmic reticulum Ca⁺⁺ pump activity and thus, plays an important role in the contraction and relaxation of cardiac and slow skeletal muscles (24-27). Although the expression pattern of phospholamban in developing hearts has been recently reported in paraffin-sections by in situ hybridization (26,27), its expression in developing AV canal region remained unclear. The cloned cDNA fragment 15H16 was obtained from RNAs isolated from stage 15 AV canal region by differential display method, suggesting a decreasing expression of phospholamban in the AV region from the stag 15 heart to the stage 21 heart. However, the whole-mount in situ hybridization showed that phospholamban mRNA signals were stronger in the stage 21 heart (Fig 3B) than in the stage 15 heart (Fig 3A).

The overall increase in phospholamban messages between these two developing stage hearts was consistent with the previous reports by Toyofuku et al. (26) and Toyofuku and Zak (24). When spatial expression patterns of phospholamban were examined in cross-sections from the same embryos after whole-mount in situ hybridization, we detected a drastic decrease in the phospholamban expression at the AV canal myocardium and the outflow tract myocardium from stage 15 heart (Fig 3C) to stage 21 heart (Fig 3D). Northern blot analysis of total RNAs isolated from stage 15 and stage 21 hearts revealed that two bands with sizes of 1.1Kb and 0.6Kb hybridized to the 15H16 probe (Fig 4A). The steady-state levels of these two messages increased from stage 15 to stage 21. Therefore, the overall expression of phospholamban in the heart increased from stage 15 to stage 21, although its expression at the AV canal region appeared to be decreased.

Skeletal alpha-tropomyosin expression is temporally and spatially decreased in the developing hearts
The clone E13 was obtained from stage 15 RNAs using the differential display method. E13 contained 268bp upstream from the poly(A) tail of chicken skeletal alpha-tropomyosin transcript (32). In avians, cardiac alpha-tropomyosin has been shown to differ from the skeletal alpha-tropomyosin in respect to its mobility in 2D gels (28,29) and at the nucleotide sequence (30-32). As can be seen in Fig 5, skeletal alpha-tropomyosin messages were detected in developing somites (open arrows) and hearts (arrowheads) at both stage 15 and stage 21 embryos. The hybridization signal was strong in the stage 15 heart and became weaker in the stage 21 heart. On the other hand, the signals became much stronger in the stage 21 somites ( Fig 5A and 5B ).

Examination of cross sections of the heart of the same embryos revealed that E13 was evenly expressed in the myocardium of stage 15 heart (Fig 5C). As the development proceeded, the expression of E13 in the myocardium was gradually diminished and at the stage 21 embryo, only the atrial myocardium expressed the E13 gene (Fig 5D). By Northern blot analysis, the E13 probe detected a 1.3Kb message in the total RNAs derived from both staged hearts (Fig 4B). Consistently, the steady-state level of E13 message was decreased from stage 15 heart to stage 21 heart. However, no change in the level of message of glyceraldehyed-3-phosphate dehydrogenase (GAPDH) was detected by Northern blot analysis of RNA derived from stage 15 and stage 21 hearts (Fig 4C). Therefore, skeletal alpha-tropomyosin recognized by the E13 probe was decreased in the developing heart temporally and spatially.

Expression of a novel 21C gene is increased at AV canal region of developing hearts
From the differential display and whole-mount in situ hybridization screening, one (21C) of the several novel genes isolated showed a cardiac-specific expression pattern during early chicken embryonic development. Whole-mount in situ hybridization using the 21C riboprobe revealed hybridization signals in the hearts of both stage 15 and stage 21 embryos(Fig 6A and 6B).

Sectioning through these embryos after whole-mount in situ hybridization revealed that the 21C gene expression was confined to myocardium (Fig 6C and 6D). At stage 15, a strong signal was found in myocardia of atrium and outflow tract ( Fig 6C ). As the development advanced to stage 21, a strong hybridization signal was detected by the 21C riboprobe in the ventricular myocardium (Fig 6B and section data not shown). The AV canal myocardium from stage 21 heart showed an increase in the expression of this 21C gene as compared to that from stage 15 heart. This is consistent with the isolation of the 21C clone from the stage 21 RNA sample using the differential display method.

To further examine the expression pattern of the 21C gene in early developing heart, we performed whole-mount in situ hybridization on stage 8⁺ and stage 13 embryos. At stage 8⁺, the 21C gene was exclusively expressed in the initially paired primordia located on either side of the embryonic midline (Fig 7A).Careful sectioning through the embryo demonstrated that 21C expression was restricted to the cardiac mesoderm and not found in any other tissues (Fig 7B). This cardiac-specific expression pattern was still detected in stage 13 embryo (Fig 7C and 7D) in which the heart tube had formed and looped to the right side of the embryo.

Northern blot analysis of total RNAs isolated from stage 15 and stage 25 hearts suggested that the 21C gene encodes a single 9.5Kb message (Fig 8B). It is also clear that the steady-state level of the 21C expression in stage 25 heart is much higher than that in the stage 15 heart (Fig 8B). This result together with the in situ hybridization data suggest an increased expression in the heart as the development of embryos proceeded. However, our in situ data do not allow us to distinguish whether more cells at the AV regions of stage 21 heart express the 21C gene or each cell at this region expresses more 21C gene or both. In a preliminary experiment, we found that the 21C expression diminished in the day 12 embryonic heart (data not shown).

For all cloned cDNA fragments used for making anti-sense riboprobes, sense probes were also made and used as negative controls. One example is shown in Fig 9, where the 21C sense riboprobe was used. This probe did not hybridize to either stage 15 (Fig 9A) or stage 21 (Fig 9B) embryos. Same results were obtained from 15H16 and E13 sense probes (data not shown).