I think it is clear what is going on once you do a few things.
Download and read the header of the PDB file for 3odu:
HEADER SIGNALING PROTEIN, HYDROLASE 11-AUG-10 3ODU
TITLE THE 2.5 A STRUCTURE OF THE CXCR4 CHEMOKINE RECEPTOR IN COMPLEX WITH
TITLE 2 SMALL MOLECULE ANTAGONIST IT1T
COMPND MOL_ID: 1;
COMPND 2 MOLECULE: C-X-C CHEMOKINE RECEPTOR TYPE 4, LYSOZYME CHIMERA;
COMPND 3 CHAIN: A, B;
COMPND 4 FRAGMENT: CXCR4 RESIDUES 2-229, LYSOZYME RESIDUES 1002-1161, CXCR4
COMPND 5 RESIDUES 230-319;
Note the keyword 'CHIMERA' and how it says 'FRAGMENT: CXCR4 RESIDUES 2-229' and then LYSOZYME RESIDUES 1002-1161 and then RESIDUES 230-319.
Next, look at 3odu at the PDB and click on the sequence tab. Usually the unmodeled line in the view under the sequence line will give you an idea for biological examples. Here it is way more complex as it is not biological and is a documented chimera and so I just think they (PDB) leave it out. Or author's didn't provide, or at least in a manner that insures the details get clearly propagated in the unmodeled line. In fact, historically PDBsum always tends to be more informative for this type of thing even though lately the unmodeled line has been better than PDB details about gaps in otherwise 'solved' structures in the past. Here is an example, see the 'Protein' tab view for the entry for 3odu at PDBsum. Note the jump in the numbering from 225 to 900 right around the SK you highlight:
I'm not going to provide an image from PDBsum for it, but again look at 1157 back to 235 a few lines below in the schematic. I bet that corresponds to near the second region you highlight. Keep in mind that whatever you did between starting with the PDB file for 3odu and now messed up the numbering so you need to compare the specific sequence.
For the exact details, you'd need to look into the methods of making this chimeric protein to see if there are linkers there aren't including in the sequence they deposited that makes up the extra few residues to connect things. Or maybe some other thing is comprising the connection represented in the gap. I'll leave looking into that to you, especially since much of what I'm seeing is stuff you should have known/expected from reading the paper.
More reading on this topic can be found here at Proteopedia's page about Unusual Sequence Numbering. Usually this sort of thing with gaps present in a structure is indicative of Missing Residues in the structures of biological chains. So be sure to read that section where it discusses how an otherwise 'solved' structure can contain regions that are not able to be models for various reasons. That page also covers the OXT issue. Eric Martz's coverage of missing residues in FirstGlance at Jmol may also help you better understand what you are dealing with and how you need to handle the related case for understanding your particular structure.
You are making this hard on anyone trying to help you.