Since the industrial revolution, global atmospheric CO₂ concentrations have rapidly increased, rising from 280 ppm to currently exceed 400 ppm (Canadell et al., 2007; Tans and Keeling, 2016). Predictions warn that the global CO₂ concentration will continue to rise due in part to humanity's continued carbon emissions (Meehl et al., 2007). The resulting increase in CO₂ will lead to a variety of both positive and negative effects on major agricultural crops used to feed the global population, many of which may yet be unknown. Elevated CO₂ concentrations, written henceforth as e[CO₂], cause increased photosynthesis in plants, which subsequently lead to positive effects such as greater growth, above-ground biomass, and yield (Ainsworth and Long, 2005; van der Kooi et al., 2016). However, e[CO₂] also causes negative effects which could have serious consequences for the quality of the crop species, such as, declines in a variety of nutrients including protein concentrations of food crops (Fernando et al., 2015; Broberg et al., 2017), vitamins and some macro- and micro-elements (Högy and Fangmeier, 2008; Myers et al., 2014). Due to these negative effects, understanding plant responses to e[CO₂] will become increasingly important as CO₂ levels rise.

The increase in photosynthesis caused by e[CO₂] results in an increase in carbohydrate production, which alters the plant's carbon and nitrogen metabolism. Apart from this direct effect on photosynthesis, many physiological processes are regulated indirectly, particularly via sugar sensing and signaling pathways. Sugar sensing and signaling plays an important role in the plant response to e[CO₂], however, this is not well understood in relation to plant nutritional quality. Sugars are well known for their use as a source of energy and organic building blocks, and in plants they also play a role in regulating gene expression (Price et al., 2004), germination (Dekkers et al., 2004), and hormonal crosstalk (Mishra et al., 2009) among other functions.

Plant growth and development requires the uptake of soil nutrients by the roots, however, the concentration of nutrients in soil can vary and plants must adapt to the environment in order to fulfill their nutrient requirements. Sugars produced from photosynthesis are transported into roots where they can assist in regulating nutrient uptake via sugar sensing (Camañes et al., 2007; Lejay et al., 2008), though little research has been done in this area. How e[CO₂] affects root function is not entirely understood, but we do know that it can affect the acquisition of soil nutrients (Taub and Wang, 2008; Pandey et al., 2015; Jayawardena et al., 2017). To what extent sugars may play a role in this is not currently known. This review aims to provide the current knowledge and understanding of sugar sensing in roots as well as the limited information available on how this is affected by e[CO₂] in order to facilitate research into this area and safeguard crops from potential negative effects of future [CO₂].

In order to study the effects of e[CO₂] in the field, free-air CO₂ enrichment (FACE) facilities have been established which allow plants to be grown in large scale open air environments. Utilizing either FACE or chamber experiments can affect the outcome of the experiment. For example, in comparison to FACE experiments, chamber studies using e[CO₂] have been shown to further increase the yield of globally important food crops (Ainsworth et al., 2008). Plant growth differences between FACE and chamber experiments are likely influenced by the root growth, as restricting the available area for root growth reduces plant biomass (Poorter et al., 2012). Most of the studies discussed in this review were conducted with chamber experiments and to our knowledge no experiments have currently been done in FACE facilities for sugar sensing studies. As such, it is uncertain how the results of many of these sugar sensing studies will potentially change in plants grown in field conditions.

Many reviews have focused on various aspects of sugar sensing, however, this review discusses the limited amount of literature published on sugar signaling and sensing as it relates to plant root function, nutrient acquisition, and hormone crosstalk. As such, we have chosen roots as the focus of our review due to the current absence of reviews in this area, but more importantly due to their importance in determining the nutrient profile of plants. This review also discusses the effect of e[CO₂] on the content of sugars in plants, including how photosynthesis and carbohydrate partitioning is affected, and how e[CO₂] may affect sugar sensing in roots. The aim of this review is to provide the information necessary for scientists developing research projects involving sugar sensing in roots or the effect of e[CO₂] on roots and sugar sensing.

Photosynthesis is a crucial process for controlling variables of crop growth and exposing C3 plants to e[CO₂] generally increases photosynthesis (Drake et al., 1997; Ainsworth and Long, 2005; Wang et al., 2012; Figure 1). Increased photosynthesis under e[CO₂] mainly occurs due to an increase in ribulose-1,5-bisphosphate (RuBP) carboxylase/oxygenase (Rubisco) activity. Rubisco catalyzes the carboxylation of RuBP, which is required for CO₂ fixation, but also uses O₂ as a substrate to oxygenate RuBP in a process called photorespiration (Makino and Mae, 1999). The carboxylation reaction of RuBP is not saturated at the current atmospheric [CO₂], therefore, as the availability of CO₂ increases under e[CO₂] conditions so too will the rate of carboxylation (Drake et al., 1997). The other process, photorespiration, is wasteful in terms of energy, as it costs the plant more energy and does not lead to any gains in energy or carbon (Peterhansel et al., 2010). However, increasing the atmospheric CO₂ levels increases the [CO₂] surrounding Rubisco, shifting the ratio of CO₂:O₂ and thereby increasing the rate of carboxylation while decreasing the rate of oxygenation (Makino and Mae, 1999).

Despite the initial stimulation of photosynthesis seen at e[CO₂], under long-term exposure to e[CO₂] the plant incurs a down-regulation of photosynthesis in both FACE studies (Ainsworth and Long, 2005) and chamber experiments (Warren et al., 2014). This occurrence is known as photosynthetic acclimation. Photosynthetic acclimation, however, does not always completely negate the positive effects e[CO₂] has on photosynthesis. For example, in one study white clover was grown under elevated (600 ppm) [CO₂] for 8 years and retained a 37% increase in photosynthesis after acclimation was observed (Ainsworth et al., 2003). These findings suggest that final growth response to e[CO₂] is largely determined by the magnitude of plant acclimation to e[CO₂].

Various explanations as to the cause of photosynthetic acclimation have been made. Decreased leaf nitrogen (N) is one such explanation. In a study on rice, e[CO₂] caused a decline in N allocation into leaf blades, which subsequently reduced Rubisco and other protein synthesis (Seneweera et al., 2011). In support of this, in a 12 year study on Liquidambar styraciflua no acclimation response occurred during the time when leaf N was sufficient for photosynthetic requirements (Warren et al., 2014). Without sufficient N to invest in Rubisco, the photosynthetic capacity of the leaf declines. Low availability of soil nitrate increases the severity of photosynthetic acclimation and seems also to be associated with an inhibition of leaf nitrate assimilation (Vicente et al., 2016). Inhibition of leaf nitrate assimilation also occurs under e[CO₂] (Bloom et al., 2014). It is not known whether the reduction of Rubisco synthesis at e[CO₂] is directly related to lower N assimilation or if Rubisco is just regulated to balance the source and sink activity.

Another explanation for plant acclimation to e[CO₂] is that an increase in sugar production tips the source—sink balance of plants, potentially resulting in more sugars being produced in source tissues than can be utilized in sink tissues. This was the case in a FACE experiment by Ainsworth et al. (2004), who used single gene mutations to test the hypothesis that photosynthetic acclimation is due to inadequate sink capacity. In the study, a soybean cultivar with an indeterminate growth trait (Williams) was compared with a line mutated for determinate growth (Williams-dt1). Only the determinate line showed photosynthetic acclimation. On the other hand, mutation of a determinate soybean cultivar (Elf) to an indeterminate form showed no increased photosynthesis. While this may provide evidence for single gene mutations being responsible for photosynthetic acclimation, this could also be explained by the fact that Elf is a cultivar bred to avoid sink limitations (Ainsworth et al., 2004). While sink capacity remains high, plants are able to continue to utilize the greater CO₂ availability. However, with limited carbon sink capacity the plant must decrease photosynthesis in order to maintain source activity. As such, when e[CO₂] causes photosynthesis to surpass what the plant is capable of utilizing or exporting to sinks, an accumulation of non-structural carbohydrates (NSC) occurs (Ainsworth et al., 2004) and leads to feedback inhibition of photosynthesis (Figure 1).

These NSCs are then able to affect gene transcription through their role as signaling molecules (Mishra et al., 2009; de Jong et al., 2014). As such, sugars are known to be involved in photosynthetic acclimation, whereby the extra carbohydrates produced under e[CO₂] cause a down-regulation of photosynthetic gene transcripts and suppress protein synthesis, thereby decreasing the rate of photosynthesis (Cheng et al., 1998). In this way, there is a feedback inhibition where the products of photosynthesis cause suppression of photosynthesis, leading to photosynthetic acclimation.

Rubisco, an essential enzyme in the photosynthetic pathway, is known to be decreased in leaves that have an accumulation of carbohydrates (Cheng et al., 1998; Aranjuelo et al., 2008). Despite this evidence, a study by Ludewig and Sonnewald (2000) opposed the hypothesis that accumulation of sugars leads to photosynthetic acclimation when they found that high [CO₂] caused accelerated leaf senescence in Nicotiana tabacum, leading to down-regulation of leaf photosynthetic related genes and thus accelerated leaf senescence. Only senescing leaves were found to show down-regulation of photosynthetic genes and increased sugar levels were not observed. They concluded that photosynthetic acclimation was caused by leaf senescence rather than sugar accumulation. Both Aranjuelo et al. (2008) and Cheng et al. (1998), however, reported that the down-regulation of photosynthesis occurred prior to senescence of the plants. All three studies used different plant species, which suggests that some species acclimate to e[CO₂] differently than others. Therefore, this suggests that photosynthetic acclimation has no single cause, with multiple processes each contributing to a different degree.

As discussed in the previous section, e[CO₂] causes an increase in carbohydrate production via the stimulation of photosynthesis. It has been observed that increased photosynthesis under e[CO₂] results in greater production of certain carbohydrates compared to others. The concentration of sucrose, the main product of photosynthesis, increases in all organs of pea plants exposed to e[CO₂] in growth chambers, however, glucose concentrations are largely unaltered (Aranjuelo et al., 2013). Glucose measurements may be inaccurate as glucose content can fluctuate throughout the day in some plants, increasing and then decreasing as the day progresses (Seneweera et al., 1995; Grimmer et al., 1999). As such, hexose to sucrose ratio will differ depending on what time period the glucose levels are measured. Glucose measurements taken when glucose levels are naturally low, will give a lower hexose to sucrose ratio than if glucose was measured during a period of high glucose levels. Sucrose levels also increased in castor oil plants grown in growth chambers under 700 ppm CO₂ compared to 350 ppm, increasing by an average of one third (Grimmer et al., 1999). Levels of sucrose are higher than that of hexoses under e[CO₂] in both chamber and field studies (Grimmer et al., 1999; Rogers et al., 2004), however, in soybean the leaf hexose-carbon to sucrose-carbon ratio increases with exposure to e[CO₂], where a five-fold greater ratio of hexose-carbon to sucrose-carbon was observed near the end of the growing season (Rogers et al., 2004). Perhaps, such variation in hexose to sucrose ratio during plant development may affect plant source and sink activities. In addition, the preference of a plant to produce one type of carbohydrate over another could potentially be linked to the control of genes by a specific carbohydrate (glucose, sucrose, etc.), though this is not known. For example, if a plant requires the presence of sucrose to initiate the repression of a specific gene, it would be ineffective to produce greater glucose quantities than sucrose. The effect that carbohydrates have on gene expression is a topic discussed further in this review, however, the impact that a change in sugar composition has on plant gene regulations is not well understood.

Starch, a major storage carbohydrate in plants, is also increased in plants growing in e[CO₂] (Aranjuelo et al., 2008). The increase in starch likely contributes to the high levels of sucrose observed with e[CO₂], due to the conversion of starch to sucrose overnight. This conversion is important for normal plant growth under ambient conditions (Smith et al., 2005), however, under e[CO₂] it may contribute to the accumulation of sucrose. In plants grown under ambient [CO₂] the starch content builds up during the day and disappears overnight. The increased production of starch under e[CO₂], however, means that not all of the plant's starch reserves are depleted during the night, leading to a gradual accumulation in leaves over time (Grimmer et al., 1999). Different plant species accumulate different amounts of sucrose compared to starch, for example spinach accumulates more sucrose and cotton more starch (Goldschmidt and Huber, 1992). These responses are likely to affect the sugar sensing pathways in either type of plant. The degree of carbon partitioning between sucrose and starch is influenced by the length of daylight. In shorter periods of light, carbon partitioning shifts toward starch synthesis, while sucrose synthesis and consumption is decreased (Pokhilko et al., 2014). Less starch is accumulated during days with long light periods, while sucrose synthesis is increased (Pokhilko et al., 2014). Sucrose content is greater during the day than night, but the amount of sucrose remaining at the end of the day, as well as the end of the night, decreases as day length decreases (Sulpice et al., 2014). The degradation of starch at night is influenced by the amount of trehalose-6-phosphate (T6P). Increased T6P was found to inhibit starch degradation at night in Arabidopsis plants, resulting in much higher starch reserves at the end of the night (Martins et al., 2013). In addition, Martins et al. (2013) found that T6P also slightly increases starch synthesis. As such, increased T6P concentrations result in more starch at both the end of the day and night. Combined with limitations on starch degradation set by the plant's circadian clock, these findings suggested a model for overnight starch metabolism (Martins et al., 2013; Lunn et al., 2014). High sucrose demand causes lower T6P, alleviating the inhibition of starch degradation and increasing sucrose content. Under low sucrose demand, T6P increases and inhibits starch degradation. The plant's circadian clock prevents the total depletion of starch at night by setting limits on starch degradation based on the length of the night period (Martins et al., 2013).

The extra carbohydrates that accumulate in leaves are allocated to the rest of the plant in varying amounts, where some organs receive more of these carbohydrates than others. Little research has been done into the allocation of carbohydrates under e[CO₂], but the following studies have investigated this. Carbohydrate allocation under e[CO₂] varies with species. Some species allocate more carbon to the seeds and others to the shoots, leaves or roots (Salsman et al., 1999; Sasaki et al., 2007; Aljazairi et al., 2014; Butterly et al., 2015). For example, during the grain filling stage of rice e[CO₂] promotes the translocation of carbohydrates stored in vegetative tissues to the panicle, as well as allocating newly fixed carbohydrates to the panicle, where it is stored as starch (Sasaki et al., 2007). A difference in carbon allocation between durum wheat and bread wheat occurs under e[CO₂]. Durum cultivars Blanqueta and Sula allocated more carbon into roots, rather than shoots (Aljazairi et al., 2014), while the bread wheat cultivar Yitpi allocated more carbon into shoots (Butterly et al., 2015). Furthermore, Sula (a modern cultivar) allocated more carbon into spikes compared to Blanqueta (a traditional cultivar), which allocated more carbon into non-reproductive shoot tissue. This indicates that variation exists within as well as between species and suggests that genetics contributes to these differences. In the case of the two durum cultivars, both differed in yield potential. Sula, which invested more carbon in spikes, is a higher yielding wheat than Blanqueta. Elevated [CO₂] also increased growth of roots and shoots of tepary bean, where the roots saw a ten-fold increase in starch (Salsman et al., 1999). Allocating more carbon into roots under e[CO₂] would contribute to greater root growth, allowing improved nutrient and/or water uptake and thus would help to maintain the balance of nutrients within the plant.

Carbon dioxide concentration is not the sole regulator of carbohydrate partitioning, with many other environmental factors involved in shaping the outcome. Which carbohydrate the increased carbon is partitioned into can be affected by the method plants use to take up nitrogen. An experiment by Aranjuelo et al. (2013) found N₂-fixing and NO3--fed plants varied greatly in sucrose content while exposed to e[CO₂]. Sucrose increased by 366% in NO3--fed plants but only by 56% in N₂-fixing plants. As e[CO₂] is known to affect the uptake and assimilation of N in plants (Bloom et al., 2014; Vicente et al., 2015a), this could point to a link between N uptake and carbohydrate allocation to roots and thereby facilitating more nutrient uptake. Plant growth method (glasshouse, field, etc.) also affects carbon allocation. Elevated [CO₂] causes increased carbon allocation to roots of perennial rye-grass resulting in increased root dry matter when grown in field conditions, however, no such results occur when grown in controlled environment chambers (Suter et al., 2002). This outcome in rye-grass was attributed to a difference in N availability, plant age and shoot sink strength. Results from Aranjuelo et al. (2013) also indicate that sink strength affects carbon allocation, where increased carbon sink strength of N₂-fixing plant's nodules allows greater storage of carbohydrates which in turn prevents the inhibition of photosynthesis by increased carbohydrates. This could mean that control of carbon allocation could be partially affected by the availability of carbon sinks. Another factor that may affect the allocation of carbohydrates under e[CO₂] is the effect e[CO₂] has on leaf area, as appeared to be the case for N allocation in rice (Makino et al., 1997). Plants which show less variable responses to leaf area under e[CO₂] (e.g., rice; Makino et al., 1997) compared to others, may allocate more carbohydrates to roots, as their leaf sink capacity doesn't change to accommodate the greater carbohydrate production. For some plants, root growth is increased under e[CO₂] (George et al., 2003), which may increase their sink capacity, allowing for greater allocation of carbohydrates to this organ. Carbon allocation under e[CO₂] can also be influenced by pH, as seen in plants grown in a low pH media under e[CO₂], where much of the carbon from photosynthesis accumulates in the shoots (Hachiya et al., 2014).

There are many reviews already written on the role of sugars as signals in plants including Granot et al. (2013), Rolland et al. (2006), and Sheen (2014) to name a few. However, to the best of our knowledge there are no reviews written specifically for sugar sensing in roots, which is a major focus of this review. As such, before moving on to our discussion of sugar sensing in roots, this section will serve to provide general information on sugar sensing not specific to roots. There is much more information known on sugar sensing than written in this section, however, we direct you to other reviews, such as those mentioned above, for more detailed discussions on sugar sensing not specific to roots.

Glucose has long been known to play a role in photosynthetic gene repression, with the enzyme hexokinase acting as a sensor (Jang and Sheen, 1994). It has since been established that hexokinase is a central enzyme in glucose sugar signaling pathways (Moore et al., 2003). Through sugar sensing, hexokinase appears to be able to promote plant growth by causing greater cell expansion in roots, leaves, and inflorescences when exposed to high light conditions (Moore et al., 2003).

In addition to hexokinase, SnRK1 has been indicated as another sugar sensor which is involved in a sucrose/T6P signaling network and operates as a starvation response (Baena-Gonzalez et al., 2007). It has been observed that SnRK1 may be inhibited by the presence of sucrose. KIN10, a part of the SnRK1 complex, is activated under sugar starvation, leading to up-regulation and down-regulation of various genes (Baena-Gonzalez et al., 2007). SnRK1 also contributes to increasing sugar content in plants by phosphorylating both sucrose phosphate synthase (SPS) and trehalose-phosphate synthase (TPS; Nukarinen et al., 2016), of which the resulting sugars, sucrose and T6P, may lead to inactivation of SnRK1 (Baena-Gonzalez et al., 2007; Zhang et al., 2009). Sucrose concentrations are linked with T6P levels, as increased sucrose leads to stimulation of TPS which in turn increases T6P concentrations (Yadav et al., 2014). High T6P then causes a decline in sucrose content which prevents further increases in T6P (Yadav et al., 2014). The regulation of T6P content is primarily linked with sucrose content, as studies have shown that only sucrose and hexoses able to be converted to sucrose have a significant effect on T6P levels (Lunn et al., 2006; Yadav et al., 2014). Sucrose and T6P may also be involved together with nitrogen assimilation, where increases in T6P signal the plant to synthesize organic and amino acids rather than sucrose (Figueroa et al., 2016). In conjunction with T6P other similar sugar phosphates, glucose 1-phosphate (G1P) and glucose 6-phosphate, are able to inhibit SnRK1, with G1P working together with T6P to significantly increase this inhibition (Nunes et al., 2013). Altogether SnRK1 appears to be involved in the plant's starvation response, inactivating during times of sufficient sucrose/T6P and activating when these signals are low.

Sugar signaling in plants begins as early as seed development and germination. At low levels, sugars are able to delay germination of Arabidopsis thaliana seeds. Other sugars have displayed this function as well, with sucrose, glucose, and the non-metabolically active glucose analog 3-O-methyl glucose exhibiting a greater delay on germination than others (Dekkers et al., 2004). The ability of the glucose analog to delay germination indicates a pathway independent of hexokinase.

Sucrose functions as a signaling molecule in a variety of ways. It is capable of inducing gene expression, such as, the Citrus ammonium transporter gene CitAMT1 (Camañes et al., 2007), as well as affecting the cell cycle. During the G1 phase of the cell cycle, sucrose induces the expression of the two CycD cyclins Cyc2 and Cyc3, which influence cell cycle progression and cell division (Riou-Khamlichi et al., 2000). The role of sucrose in regulating the cell cycle likely correlates with its role in plant growth. As a plant produces more sugars, sucrose stimulates the cell cycle and allows utilization of the produced sugars for growth. As such, e[CO₂] is likely to facilitate this process. The greater sugar production caused by e[CO₂] could stimulate the cell cycle and allow the excess sugars to be used to produce greater plant biomass (Seneweera and Conroy, 2005).

Sugar signaling pathways also interact with hormones. For example, glucose increases the biosynthesis of auxin, therefore affecting processes regulated by this hormone (Sairanen et al., 2012). Evidence also suggests that sugars interact with pathways of both abscisic acid (Cheng et al., 2002) and ethylene (Price et al., 2004). Among other functions, abscisic acid has an enhancing effect on some genes regulated by sugar (Rook et al., 2001), while glucose downregulates the expression of ethylene biosynthetic genes (VnACO2 and VnEIL1) and a transcription factor involved in the ethylene signaling pathway of narbon bean cotyledons (Andriunas et al., 2011). These findings show the various roles of sugars in gene regulation and thus their contribution to plant growth and development by way of sugar sensing.

Currently there is a lack of understanding about the effect of e[CO₂] on sugar sensing, however, many studies have conducted experiments applying exogenous carbohydrates to plant roots, thus creating conditions of increased root sugar content which may mirror the conditions of greater root sugar content resulting from increased photosynthesis under e[CO₂]. Most of the research into the role carbohydrates play in plant roots has focussed on sucrose exclusively. While some research has brought to light several effects of other carbohydrates, such as, glucose and fructose, there may yet be many more roles that non-sucrose carbohydrates play. Much of this work is limited to A. thaliana, but it is likely that sugars play many other diverse roles in root function that may be discovered among other plant species. The following section discusses the potential outcomes for roots of plants grown under e[CO₂], whereby excess carbohydrates in leaves are transported to roots and lead to altered gene expression (Figure 1). The effects of sugar sensing in roots has had less attention then in shoots, as is especially the case for sugar sensing under e[CO₂]. As such, there is insufficient data to draw conclusions at this time, however, we provide an insight into how e[CO₂] may affect sugar sensing in roots, as well as sugar crosstalk with hormones.

Much is still unknown about how plants will react to e[CO₂] and with nutrient deficiencies observed in agricultural crops, this will become increasingly more important to understand. The production of carbohydrates is increased in plants grown under e[CO₂] due to an increase in photosynthesis. Some carbohydrates are produced in higher quantities than others depending on the plant, though production of sucrose is reportedly higher compared to hexoses. The studies discussed provide an insight into how these sugars can be used to regulate many functions in roots. Most of the information on sugar signaling discusses the glucose and sucrose pathways. The amount of carbon partitioned into either of those carbohydrates may be in part determined by which carbohydrate the plant requires to regulate specific genes, though this is unknown. Nutrient acquisition appears to be regulated by sugars, as evidenced by the regulation of expression of various ion transporters as well as the ability for sugars to affect root growth. Finally, both e[CO₂] and sugars are able to affect the biosynthesis of certain plant hormones, which may suggest that sugars function as an intermediate in e[CO₂] control of hormones. From these studies we can begin to think about what changes might occur in roots of plants grown in future carbon dioxide concentrations.

MT wrote the manuscript. SS, AM, NH, and DG each contributed to the design of the review as well as revision of drafts and the final manuscript.