These are not artificial chromosomes but they are what we are aiming to make. Chromosomes which exist in the cell nucleus as independent DNA molecules and which are segregated efficiently at cell division and transmitted through meiosis into the gametes of mammalian cells. The chromosomes shown here are mouse metaphase chromosomes which have been visualised by scanning electron microscopy (from Adrian Sumner). The reasons for our interest are sumarised in the next two slides
Two fundamentally different routes have been taken to end up with modified chromosomes. In the first the start point is an existing chromosome which is reduced in size by a variety of methods. These are engineered chromosomes
In the second approach defined DNA components representing telomeres and centromeres are assembled to create a new chromosome termed a Mammalian Artificial Chromosome (MAC) or Human artificial chromosome (HAC).
To take this last approach we had to start from the information available on the functionality of telomeres transfected into cells and take the best available guess one what sequences might constitute the centromere.
Mammalian centromeres (or kinetochores as the assembly of DNA and proteins is termed) are large complex organelles the with many functions required to ensure segregation of the daughter chromosomes at mitosis. These include checkpoint machinery moleular motors and probably many functions as yet unknown.
Sequenced Mammalian Centromeres?
In some organisms the centromeric DNA has been sequenced but in mammals the kinetochore is usually embedded in a block of heterochromatin containing highly repetitive DNA. Current large scale sequencing technologies are not designed to deal with these regions of the genome.
The nature of the centromeric sequences are a matter of controversy. An argument have been made that the sequence is a small "magic" one which is evident in the simple chromosomes of S. cerevisieae but which is hidden by the repetitive sequences in mammalian chromosomes.
Point centormer of S cerevisiae - a magic sequence
A contrasting argument is that in some instances new centromeres have been observed to form on previously non centromeric DNA after chromosome rearrangements. In one case analysis at the sequence level shows no obvious features and no changes in methylation as a result of the change form non centromeric to centromeric function.
The best bet sees to be to use the repetitive sequences themselves of which in human chromosomes alphoid DNA, present at all centromeres is the strongest candidate.
Two similar methods based on assembling alphoid DNA and telomeres have been successful. In one multimers of an alphoid repeat unit were generated, cloned and mixed with telomeric sequences and genomic DNA prior to lipofection into a cell line. In the second vertebrate telomeres were recombined onto a yeast artificial chromosome and the DNA from this lipofected or micro-injected into mammalian cells.
Methods of making a chromosome
Alphoid DNA is variable in structure, some blocks of the sequence are regular arrays of a monomer, some blocks have diverged and irregular arrays, some contain the binding site for a CENPb protein and some, particularly on the Y chromosome, do not. Some of these features may be more important for forming a centromere/kinetochore than for replicating and already formed one.
The opposite approach of fragmenting and existing chromosome with telomeres has been used most extensively on the human X and Y chromosomes. Passage of one of these types of chromosomes through different cell types by microcell fusion produces rearranged chromosomes.
Top down chromsome fragmentation
In the case of one chromosome produced like this de novo centromere formation seems to have taken place as the chromosome has no mouse minor satellite DNA or human alphoid DNA
Chromosome 310 with different probes
Combined in situ hybridisation and immunocytochemistry suggests that the Y chromosome long arm sequence are providing the DNA underlying the kinetochore in this chromosome.
Fiber fish - co localisation of centromere antigens and Y sequences
The initial technology used to make "bottom up" chromosomes was complex – we have addressed the following three questions
By using PAC vectors, which are propagated in E. coli, DNA production is simplified and the MAC construct can be produced in forms with and without telomeric DNA and in linear and circular versions.
Circular constructs based on alphoid DNA from chromosome 21 function efficiently in forming MACs
Here is an example of such a chromosome
A chromosome made from circular DNA
Telomeric sequences are not necessary when circular molecules are used
The kinetochores seem to be normal based on the presence of the expected centromeric proteins
Centromeric proteins are located on these kinetochores
Unless genes can be expressed from these chromosomes they will be of little practical use. The presence of large amounts of repetitive sequences might be expected to set up heterochromatic regions on the chromosome, which might silence gene expression. To test this we have constructed a chromosome containing the genomic HPRT locus by co-transformation of HPRT- HT1080 cells.
Chromosome made form alphoid DNA and HPRT PACs
A mac containing the HPRT locus (this is the link)
This MAC is stably propagated during cell division
HPRT mRNA can be detected in this cell line after many generations growth in the absence of selection for the marker genes
Northern blot of HPRT expression
Bottom up approach - positives and negatives