$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
The present article aims to detail the step by step protocol used to purify chloroplasts (and their sub-compartments) from Arabidopsis thaliana. Since the availability of its complete genome sequence almost two decades ago, and of large collections of insertion mutants made available to the community, Arabidopsis is now widely accepted as a model plant. However, while this plant was perfectly adapted for genetic approaches, plant scientists needed to adapt biochemical and physiological tools to this emerging model. Protocols allowing to purify photosynthetically active chloroplasts from leaves of well-established biochemical models like spinach12 or pea13 thus had to be adapted. The first method describing purification of Arabidopsis chloroplasts was published in 199814, just before the release of the Arabidopsis genome sequence. Several years later, simple methods for isolating Arabidopsis chloroplasts compatible with studies aiming to analyze in vitro import of proteins in purified organelles were made available15,16. However, these methods did not allow to combine high level of purity and preservation of photosynthetic activity of the purified chloroplasts. More recently17, a rapid method was established, which relies on the use of Percoll gradients, and allows to retain almost 90% of the photosynthesis rate measured in the starting leaves of Arabidopsis.
The protocol described here allows to purify Arabidopsis chloroplasts at an excellent level of purity. Indeed, immunological detection of contaminants from other cell compartments demonstrated that the purified organelles are devoid of mitochondrial and plasma membrane markers9,10. This protocol was also efficient to purify chloroplasts from several Arabidopsis ecotypes18, like Columbia (Col) or Wassilewskija (WS), i.e., the ecotypes that were used for genome or expressed sequence tags (ESTs) sequencing projects but also to generate T-DNA insertion mutants in Arabidopsis. In other words, when proteomics studies have to be performed, the present protocol is compatible with these two reference ecotypes from Arabidopsis. Finally, the yield of chloroplasts using the present protocol is similar to the one obtained when starting from spinach or pea leaves (i.e., 3%, as measured from the chlorophyll content in the Percoll-purified chloroplast when compared to the total chlorophyll amount present in starting leaves). In terms of proteins, the yield is close to 50 mg of chloroplast proteins, when organelles are purified from 500 g of 5-week-old Arabidopsis leaves.
To reach such a good yield (and chloroplast integrity), one should however pay special attention to several steps when using the present protocol. The chloroplast in Arabidopsis is an extremely fragile structure (this is not the case for pea chloroplasts, for example). Specific attention is thus required in order to avoid large-scale rupture of the organelles during purification. The number and size of starch granules present in chloroplasts are critical for the preparation of intact chloroplasts. Indeed, chloroplasts containing large starch grain will generally be broken during initial differential centrifugations steps aiming to concentrate the crude chloroplast fractions12. Therefore, the plants should be kept overnight in a dark and cold room (4 °C) prior to the experiment, to reduce the amount of starch.
New users of the present protocol could be tempted to start from larger amounts of leaf material (huge rosettes from old Arabidopsis plants with larger leaves) trying to enhance the recovery of purified chloroplasts. However, in our hands, starting from young leaves (5-week-old) is the best compromise to combine yield, purity, and integrity of the purified organelles. Indeed, too old leaves are highly enriched in phenolic compounds that were shown to have a negative impact on chloroplast integrity19.
Finally, the initial extraction step (grinding of the tissue) is another critical step. The blending process must be limited to few seconds. As stated above, new users might be tempted to use longer blending, thus expecting to strongly improve the yield of purified organelles. However, if longer blending effectively releases more material from leaves, it appears that the proportion of broken chloroplasts rapidly increases in the crude cell extract. Due to this high ratio of broken to intact chloroplasts in the medium, further purification steps (separation on Percoll gradients) are strongly affected and the yield of the purification is unexpectedly lower.
Availability of specific protocols to purify organelles have allowed a series of high throughput proteomics-based experiments to be conducted on chloroplast samples. These data were made available in several public databases6, thus providing to biologists in the field an accurate subcellular (and subplastidial) localization for many chloroplast proteins. This was especially true for envelope proteins whose identity and location remained mostly unknown before these analyses, since envelope membranes represent a minor chloroplast component (1-2% of the chloroplast proteins) while playing a key role in chloroplast metabolism and biogenesis5,20. Using the protocol described here, we recently analyzed the composition of the three main chloroplast compartments from Arabidopsis (i.e., the stroma, the thylakoids, and the envelope membrane system)9. Based on a semi-quantitative proteomics approach (spectral counting), we were able to assess the partitioning of hundreds of proteins in these three chloroplast compartments.
While the present protocol allows to purify the three main compartments of the chloroplast from Arabidopsis, it is also possible to distinguish additional sub-compartments in the chloroplast. Indeed, the envelope membrane system is made of the inner and the outer envelope membranes (Figure 1). However, to the best of our knowledge, a method to purify inner and outer envelope membranes from Arabidopsis chloroplasts remains to be established. Inner and outer envelope membranes can be purified from spinach21 or pea22 chloroplasts. The main limitation of Arabidopsis mostly results from the limiting amounts of starting material. Starting from 500 g of Arabidopsis leaves (which already requires 1 m2 surface in a growth chamber) allows purifying only 100 µg of envelope proteins. On the other hand, it is easy to start with 5-10 kg of spinach leaves from the market, to purify large amount of chloroplasts8 and to end with a yield of 3 to 10 mg of envelope proteins from this material.
The same is true for thylakoid sub-compartments. Indeed, thylakoids are made of light membrane vesicles (lamellae) and dense structures (grana) (Figure 1). Specific protocols are available to distinguish these two compartments in Arabidopsis23,24. Again, based on a quantitative proteomics analysis, we recently inventoried the proteins present in these two sub-compartments24. These approaches, together with an in-depth investigation of the literature, allowed validating, or proposing hypotheses for, the subplastidial location of hundreds of thylakoid proteins. However, it is important to note that additional membrane microdomains are present at the curved margins of thylakoids. These lipoprotein sub-compartments, or plastoglobules, are permanently coupled to thylakoid membranes and contain a specific set of proteins25. Using the present protocol, it is thus not possible to distinguish these specific proteins from other thylakoid components.
Some genuine (well-known) envelope, stroma, or thylakoid components are still lacking from the lists of detected proteins. Together with targeted biochemical and immunological analyses, the continual improvement of MS sensitivity will be of great help to revisit the chloroplast content towards a complete repertoire of the composition of its various sub-compartments.