Introduction:  Within a eukaryotic cell, there are a number of different, and physically discrete compartments: internal organelles (mitochondria, peroxisomes, chloroplasts -- not shown), the secretory system (endoplasmic reticulum, Golgi apparatus, etc), and the nucleus.

The compartmentalization of enzymatic processes has distinct advantages, it allows what might be conflicting or incompatible reactions to occur within a single cell by restricting them to physically separate compartments.

In addition, within a particular compartment, specific enzymes can be placed together. This co-localization increases the efficiency of chemical reaction systems, such that the product of one reaction is generated in the immediate vicinity of the enzymes that use that molecule as a substrate for subsequent reactions.

Such juxtapositioning of enzymes can be particularly critical when the by-product of a reaction is toxic, and must be destroyed or transformed rapidly in order to protect the cell itself, or the organism as a whole.

Compartmentalization increases the efficiency of chemical reactions, and is therefore subject to selective pressures. Compartmentalization is the major structural difference between us (eukaryotes) and bacteria.

You can learn more about intracellular organelles at

Typically, specific proteins are found within the same subcellular compartment in all related organisms. For example, the proteins of the electron transport chain -- responsible for ATP synthesis -- are located in the inner mitochondrial membrane in all eukaryotic cells, while many hydrolases, associated with the breakdown of ingested materials, are located in lysosomes.

For a protein to be localized to different organelles in different related organisms is rare, and presumably must be driven by specific evolutionary selection pressures. Perhaps the most dramatic known example of altered intracellular localization involves the enzyme alanine:glyoxylate aminotransferase 1 (AGT).

Among vertebrate species AGT can be found in the cytoplasm (e.g. in guinea pigs & frogs), within the peroxisome (in gorillas, orang-utans, baboons, rabbits, guinea pigs, fruit bats, wallabies and koalas), within mitochondria (cats, dogs, ferrets, moles hedgehogs, frogs, newts and turtles), or evenly divided between peroxisomes and mitochondria (e.g. in marmosets, tamarins, lemurs, rat, mice, squirrels, porcupines).

In humans, AGT is normally a peroxisomal enzyme. However, in a subtype of the genetic disease "primary hyperoxaluria type I" (PH1), AGT is found localized to mitochondria. This mislocalization can lead to kidney failure and death. In severe cases, kidney transplantation is required to rescue the patient.

How AGT comes to be localized to mitochondria in these PH1 patients is due to a combination of genetic defect and selective pressure.

Targeting signals:  Within a eukaryotic cell, RNAs are synthesized in either within the nucleus or within mitochondria or chloroplasts. In the case of nuclear mRNAs, the primary transcript must be processed (spliced, capped and polyadenylated) and then transported from the nucleus to the cytoplasm. In the cytoplasm, the mRNA engages the translation machinery and begins its translation into a polypeptide.

Polypeptides that are fated for secretion or for insertion into one of the cell's membraneous systems (i.e. endoplasmic reticulum, Golgi apparatus, transport vesicles, plasma membrane), are targeted to the endoplasmic reticulum by the presence of an N-terminal "signal sequence". As the signal sequence emerges from the ribosome, it is recognized by the "signal recognition particle", and the SRP/Ribosome/nascent polypeptide complex is delivered to specific transport complexes located on the ER surface. Once docked with the ER the nascent polypeptide is threaded through the transport machinery, to emerge on the lumenal side of the ER membrane. At this point the signal sequence is often removed by proteolysis, and the polypeptide is either released into the ER lumen, or anchors into the ER membrane. Within the ER other "signals", embedded in the polypeptide's sequence, are used to direct the polypeptide to either remain in (or be returned to) the ER, or move onto the Golgi apparatus, lysosome, plasma membrane or to be secreted.

Proteins not directed for the secretory/membrane system are made on non-membrane bound "free" ribosomes. As these polypeptides are synthesized, they undergo a process of folding (which is often facilitated by proteins known as chaperonins, and are then released from the ribosome, to float free in the cytoplasm. In some cases, the nascent polypeptide rapidly assembles into higher order structures.

Depending upon where they need to go within the cell, these nascent polypeptides have different fates.

A cytoplasmic protein is, to a first approximation, already in its appropriate place upon release from the ribosome. In some cases, positioning within the cytoplasm is influenced by where in the cell its mRNA is translated. Actin display this type of RNA-dependent subcytoplasmic localization. In most cases, however, subcellular localization appears to be based on interactions with specific structures, such as the plasma membrane.

Proteins destined for the nucleus enter through nuclear pores, located within the nuclear envelope.

In the case of mitochondrial, and peroxisomal proteins, however, there is a barrier in the form of a lipid bilayer that must be crossed in order for them to reach their final destinations. A polypeptide must pass through one (for the peroxisome) or two (for the mitochondria) lipid membranes. Most mitochondrial proteins are transcribed from nuclear genes, so they are originally translated in the cytoplasm.In contrast to secreted proteins (see above), their translation is not coupled to transport. A cytoplasmic mitochondrial polypeptide typically contains an N-terminal "leader" sequence that interacts with a "translocation complex" on the mitochondrion's cytoplasmic surface. The current view is that movement of a mitochondrial polypeptide through the translocation complex involves the unfolding of the polypeptide, and its refolding once within the mitochondrion. In the course of transport, the leader sequence is usually proteolytically removed and chaperonins, found within mitochondrion, facilitate its refolding. That the unfolding of transported protein is critical for transport is suggested by the fact that tightly folded polypeptides, coupled to a mitochondrial leader sequence, associate with the mitochondria surface but are not transported, rather they seem to "plug up" the transport machinery. Mitochondrial transport can be blocked if the unfolding of the cytoplasmic form of the polypeptide is energetically unfavorable.

Targeting pathways for AGT, taken from C. Danpure 1997. Variable peroxisomal and mitochondrial targeting of alanine"glyoxylate aminotransferase in mammalian evolution and disease. BioEssays 19:317-326.

In contrast to the case with mitochondrial polypeptides, polypeptides destined for the peroxisome do not appear to be unfolded during transport -- in fact gold particles of up to 9 nm in diameter associated with a "peroxisomal targeting sequences" (PTS) are transported into the peroxisome (Walton et al 1995), arguing that large enzymes can betransported "intact", i.e. without unfolding, into the peroxisome. Since they enter intact, there seems to be no need for chaperonins within the peroxisome and none have yet been described.

Two types of PTSs' have been identified. The type I PTS is a tri-peptide sequence located at the C-terminus of a polypeptide, whereas type 2 PTSs are located at the N-terminus. Both types of PTSs appear to interact with cytoplasmic adapter polypeptides; this complex then interacts with the protein transporter system located within the peroxisomal membrane (FIG). The similarities between peroxisomal and nuclear transport are striking, but a 'peroxisomal pore" structure has yet to be identified ultrastructurally. 

Primary hyperoxaluria type 1 (PH1):  Primary hyperoxaluria type I is a rare autosomal recessive disease. Its cause is the absence of functional AGT in peroxisomes. A number of distinct types of mutations are able to generate PH1.

In ~40% of PH1 patients, the AGT protein is absent, generally due to nonsense mutation that prematurely terminates translation, leading to an inactive and high unstable polypeptide.

In ~16% of patients the AGT protein is present but inactive, presumably due to mutations that disrupt the enzyme's active site, rendering it catalytically "dead".

In ~40% of patients AGT is present and active, but mislocalized -- it is no longer directed to peroxisomes, but is found in mitochondria! How human AGT gets back to the mitochondria is a tale of multiple molecular events and changing diets.

In those cases of the PH1 due to AGT mislocalization, the disease is caused by the combination of two genetic changes. Approximately 20% of Caucasians carry a allele of the AGT gene in which the normal proline residue at position 11 has been replaced by a Leu residue (Pro11» Leu11).

The Pro»Leu change leads to a generation of new, but relatively weak mitochondrial leader sequence. On its own, however, the Pro»Leu form of AGT is not localized to mitochondria efficiently. The reason appears to be that AGT normally dimerizes rapidly after its release from the ribosome.

Mitochondrial transport requires the polypeptide to be "unfolded" in order to be efficiently transported. Dimer formation inhibits this unfolding, and so the mitochondrial transport of AGT; peroxisomal transport, which can deal with large, folded molecular complexes, is not effected. The mitochondrial import sequence encoded by the translation 1 start site, which was lost in human due to a mutation, appears to inhibit AGT dimerization, and thereby allow efficient mitochondrial transport.

In people homozygous for the Pro»Leu allele of AGT (about 4% of a Caucasian population), ~5% of total AGT is found in mitochondria, and the rest in peroxisomes.

During mitochondrial transport, this leader sequence is proteolytically removed, allowing AGT dimerization to occur within the mitochondria. It is tempting to speculate that early in human evolution, dietary habits made the need to localize AGT to peroxisomes so critical that a mutation that removed the upstream translation start site was highly advantageous.

Later in human evolution, however, as the diet of certain populations moved from primarily herbivorous to more carnivorous, the selection pressure changed and a selective advantage was bestowed on those that could localize at least a small amount of AGT to mitochondria -- presumably based on their increased ability to form glycogen. and thereby store metabolic energy. It was this advantage that lead to the high allelic frequency of the Pro»Leu in certain populations.

Once the Pro»Leu mutation was in place, however, a second mutation was required to develop the "mitochondrial AGT" form of PH1.

A common mutation is Gly170»Arg170. On its own, this mutation does not significantly alter either the activity or intracellular localization of the AGT polypeptide.

In combination with the Pro»Leu allele, however, dimerization of AGT is inhibited. Slow cytoplasmic dimerization allows time for the Pro»Leu-dependent mitochondrial leader sequence to engage the mitochondrial transport machinery, and so greatly increases the efficiency of mitochondrial transport.

Once inside the mitochondria, AGT can form dimers and become active. The increased efficiency of mitochondrial localization of AGT would be associated with a decrease in the peroxisomal levels of AGT.

It is this decreased level of AGT that apparently leads to the formation of kidney stones, kidney damage and death in untreated individuals.

Thought questions

1. Speculate on the factors that could effect the severity of PH1 in people homozygous for the Pro»Leu/Gly»Arg allele of AGT?

2. From the perspective a gene therapy, how might you treat a "mitochondrial AGT" form of PH1? How would this approach differ from a that used to treat a person who is completely missing AGT activity?