Development of the vertebrate hindbrain starts with a series of constrictions of the neural tube. This results in the formation of eight segments known as rhombomeres. There is strong evidence that each rhombomere has a distinct identity. How rhombomeres acquire identity and which molecules are involved in this process is a largely open question. We have recently isolated a homeobox gene (Ghox-lab) that may be critical for the specification of rhombomeres. With the help of a specific antibody, Ghox-lab protein is detected in posterior neural plate well before the appearance of rhombomeres. An intricate sequence of events involving localized increases and decreases in transcription eventually results in the formation of a distinct island of expression of Ghox-lab in rhombomere 4. All cells in the rhombomere 4 express Ghox-lab, as do the neural crest cells migrating away from this rhombomere. These observations led us to hypothesize that Ghox-lab is involved in the specification of segment identity in the chick hindbrain. We now propose to more closely study the function and regulation of Ghox-lab in the hindbrain as follows: (1) The Ghox-lab gene gives rise to two transcripts. We will isolate the corresponding cDNAs and determine their spatial distribution in the embryo by in situ hybridization. We will also examine whether the striking increase in transcription in rhombomere 4 involves one or both mRNAs. (2) Ghox-lab protein is found in the neural plate up to a region in which the anterior boundary of rhombomere 4 will later form. Anterior hindbrain, mid- and forebrain do not express Ghox-lab- To examine whether cell interactions and other extracellular cues affect Ghox-lab expression, we will graft a piece of anterior neuroectoderm into the region where Ghox-lab positive and negative cells meet. The status of Ghox-lab expression in graft and host will be examined with Ghox-lab antibody. To distinguish graft and host tissue we will use chick-quail chimeras. (3) In an effort to obtain the entire Ghox-lab gene, including its 3' and 5' regulatory sequences, we will construct and screen a genomic P1 phage library that can accommodate inserts up to 100 kb. (4) To identify and study the regions in the Ghox-lab gene that are responsible for the rhombomere-4-specific pattern of expression, we will produce transgenic mice with a P1 clone that contains the entire Ghox-lab gene. We will identify Ghox-lab protein in these mice with the help of the (chick specific) Ghox-lab antibody. Once we have established the native expression pattern, we will generate transgenic mice with deletions in Ghox-lab. (5) Ghox-lab is inducible by retinoic acid (RA) in gastrula stage chick embryos. We propose to examine whether RA is required for the normal expression of Ghox-lab by determining the ability of gastrula embryos to synthesize RA from the precursor retinol.
