|Home | About | Journals | Submit | Contact Us | Français|
Eukaryotic organelles encapsulate defined subsets of cellular biochemical pathways. For example, beta oxidation of fatty acids occurs inside mitochondria while fatty acid chain elongation takes place on the endoplasmic reticulum membrane. Organelle membranes isolate reactions from each other and store intermediates and products, and can thus be viewed as “reaction vessels”, playing roles analogous to the reflux columns and holding tanks of a chemical factory. To develop an effective chemical manufacturing process, it is not enough to focus just on the chemistry, i.e. the reactants and solvents that directly participate in reactions. The size and design of the reaction vessels is of equal importance. Likewise, within a cell, the size of organelles will influence the rates of biochemical pathways contained within them. Organelle surface area can limit the rate of import of substrates and efflux of products, while the volume of the organelle can dictate the quantity of intermediates that can build up (Figure 1). Many key metabolic enzymes are organelle membrane proteins, and in such cases increased surface area could allow larger numbers of molecules into the membrane to increase metabolic flux.
The influence of organelle size on metabolism is indicated by the fact that in cells specialized for certain pathways, the organelles that contain these pathways are enlarged compared to other cell types. Secretory cells are an obvious example, in which the requirement for a high rate of flux of secreted proteins is met by a massive over proliferation of endoplasmic reticulum and Golgi apparatus. Other examples include enlarged lipid droplets in adipose cells, proliferation of microvilli on the surface of cells lining the intestine, increased surface area and volume of rhodopsin containing vesicles in rods versus cones, and changes in mitochondrial abundance as a function of respiratory state.
If, as we hypothesize, organelle size affects metabolism and signaling, then reprogramming of organelle size could be used as a novel strategy for reprogramming cellular state and behavior, with direct applications in medicine and biotechnology.
Cytopathologists diagnose cancer by visual assessment of cell geometry including organelle size. For example, enlarged nuclei in a pap test indicates early stage cervical cancer. Cytopathology texts are full of such examples, but we don’t understand why these changes occur. According to the hypothesis of this review, these changes of cell geometry in cancer arise because cells have adapted to the metabolic alterations that are a hallmark of cancer . Could we attack cancer cells by reprogramming organelle size?
We can distinguish two possible reasons for organelle size alteration in cancer cells, which in turn predict two possible ways that organelle targeted therapy could be useful (Figure 2). First, if organelle size is adjusted as a response to pathological alterations in cell metabolism, then if we could reprogram organelle size in a cancer cell using small molecules that target the size control pathway, the cell might die due to a mismatch between organelle size and metabolic state. Alternatively, organelle size alterations might arise from pathological alterations in signaling pathways that impinge on the size control system, and then alterations in cell metabolism or behavior would be a downstream effect of the change to organelle size. In this case, it might be possible to drive the cell back to a less malignant state by driving its organelles towards a more normal size range. Either outcome would be therapeutically useful, but so far this “organelle directed medicine” strategy has not to our knowledge been tested in any cancer model system.
Reprogramming organelle size would also have applications in metabolic engineering. Increasing the size of organelles that encapsulate key steps of metabolite production, especially those involving toxic intermediates, could greatly enhance metabolite production. For example biodiesel production could be enhanced by targeting genes that control lipid droplet size [2–3] thereby enhancing the ability of the cell to store triglyceride (TG).
Before we can implement or test these applications in medicine and biotechnology, we need to obtain mechanistic understanding of how organelle size is regulated.
Three distinctions arise in considering organelle size, the first being size versus number. A particular total quantity of organelle, as judged by volume and surface area, could either be manifest as a single large organelle or a collection of smaller but more numerous organelles. From a “reaction vessel” perspective the total quantity of the organelle matters more than the number of organelles into which it is partitioned.
A second important distinction is size versus scaling. Scaling describes a relation between organelle size and overall cell size . Based on our overall hypothesis that organelle size is tuned to the biochemical needs of the cell, larger cells should have larger organelles. In fact this is often observed, for example for nuclei [5–6] and mitotic spindles .
Finally, we note that the “size” of a membrane bound organelle could refer either to surface area or volume. Are both regulated, or is one regulated actively, with the other passively responding? Cellular membranes are permeable to water [8–9], so organelle volume can change just by letting water in or out [10–11]. In contrast to volume, surface area changes require active addition or removal of membrane.
In order to test the idea that organelle size affects cellular function, and to test the strategy of altering organelle size to reprogram cell behavior, we will have to develop tools to alter organelle size in a predictable way. Development of such tools must await a better mechanistic understanding of how organelle size is actually regulated. We next consider five possible classes of organelle size control systems along with suggestions for experiments to test which mechanism is at play for any given organelle.
One way to control the size of an organelle is to produce a fixed quantity of a critical structural precursor, so that the quantity of that molecule sets the size of the organelle . Larger cells have more ribosomes, producing more protein and thus a greater quantity of the limiting precursor, automatically producing organelle size scaling.
An test for the limiting precursor mechanism is that organelle size should be inversely proportional to organelle number in the cell. This test supported a limiting precursor model for centrosomes  but not flagella . Molecules that regulate expression of limiting precursors represent targets for developing small molecules to alter size. Since the molecules of which an organelle are composed are the same molecules that perform organelle functions, it is likely that the regulatory pathways that regulate production of metabolic enzymes will control limiting precursor production.
If cells and organelles each grow at constant rates, then the organelle size to cell size relationship will be characterized by a power-law scaling relation, with a slope dictated by the difference in growth rates . Constant growth of cells and organelles thus produces scaling directly without measuring organelle size or adjusting organelle growth rates over time. Enzymes involved in organelle component biosynthesis would be targets to develop small molecules to alter size in this type of mechanism.
Constant growth can be tested by ablating the organelle. If the growth rate increases to “catch up” with the pre-ablation organelle size, this would argue against the constant growth model. This test was used to reject the constant growth model for scaling of flagella  and nuclei , both of which increase their growth rate if they become too small relative to the size of the cell.
One way to regulate factors that modify organelle morphology would be for the cell to have sensors capable of measuring organelle size or shape, and then use the output of these sensors to regulate organelle assembly or disassembly. But how would a cell measure organelle size? For linear structures, ruler molecules have been proposed whose length matches the desired length of the structure [17–20]. For membrane bound organelles, ion channels in the membrane, if present at uniform density, could produce surface area dependent ion fluxes that could potentially serve as indicators of size. Such channels could serve as targets for developing molecules to reprogram size in such instances.
An alternative feedback scheme is for the cell to monitor organelle function and adjust organelle growth accordingly. For example, an organelle responsible for biosynthesis of a particular product might have its growth increased in response to a need for increased quantities of that product by the cell.
This type of feedback has been observed in regulation of endoplasmic reticulum (ER) quantity by the unfolded protein response (UPR) pathway which monitors efficacy of protein folding within the ER and adjusts ER biosynthesis accordingly [21–22]. Regulatory molecules that respond to functional output, such as components of the UPR pathway, are potential targets for developing small molecules to reprogram size.
Function-based feedback control allows organelle size to automatically scale to the physiological needs of the cell. Experimental testing of function-based feedback can be accomplished by inhibiting the normal function of the organelle and asking whether this results in increased organelle size.
The steady-state size of dynamic organelles is a function of the rates of assembly and disassembly. If either of these rates is inherently size-dependent, then it is possible to have a system in which there is only one possible steady state solution, giving size control without a need for extrinsic systems to monitory organelle size. This type of system has been implicated in eukaryotic flagella, whose axonemal microtubules undergo continuous disassembly at a constant rate which is balanced by continuous assembly, and since the assembly is dependent on transport from the cell body out to the tip of the flagella, the assembly rate is a decreasing function of length . Since the assembly rate is a constantly decreasing function of length, while the disassembly rate is constant, there is just one value for length where the two rates balance each other, giving rise to a unique steady state length .
Any change to processes that affect the stability of the structure can lead to changes in steady state size. For example alterations in microtubule turnover enzymes can lead to changes in mitotic spindle size . Alterations in vesicle fusion and budding rates may have similar effects on the size of membrane bound organelles . Enzymes involved in organelle assembly or disassembly thus provide potential targets for size-altering compounds.
Clearly, our understanding of organelle size control systems is still at a primitive level. Already, though, researchers are discovering molecular perturbations that can alter organelle size. The time is therefore ripe to begin testing whether such size perturbations can really alter cell function. If so, it will provide additional impetus for further investigation into size control systems, while opening up a whole new areas for applications, such as organelle size-directed therapy and size-directed metabolic engineering. The big missing piece for testing the hypothesis that altering organelle size can reprogram cellular function is the discovery of small molecule (or genetic) modulators size. One approach to identify such molecules or mutations is unbiased screening of large collections, but as discussed above, better understanding of the mechanisms that regulate size would automatically suggest possible targets. Furthermore, without understanding the mechanisms that control size, it will be extremely difficult to interpret the mechanism of action of any molecules or mutations that might be found by random screening. The best way to pursue the idea of organelle directed medicine, at this point, is to understand how size is regulated in the first place.
I wish to thank Mark Chan, Prachee Avasthi Crofts, and Susanne Rafelski for careful reading of the manuscript, and all members of my lab for helpful discussions. Our work on organelle size control is supported by NIH grants R01 GM097017 and P50 GM081879.